1 Can we improve the chilling tolerance of maize photosynthesis through breeding? Angela C. Burnett* Johannes Kromdijk Address for all authors: Department of Plant Sciences, University of Cambridge, Downing Street, Cambridge, CB2 3EA, UK *Author for correspondence Email addresses for all authors: acb219@cam.ac.uk jk417@cam.ac.uk Word Count: 9246 Running Title (max. 60 characters): Improving photosynthetic chilling tolerance in maize Highlight: Photosynthetic chilling tolerance is composed of several physiological mechanisms, underpinned by differing amounts of genetic variation. A holistic, high-throughput approach is needed to improve chilling tolerance of photosynthesis in maize. 2 Abstract 1 2 Chilling tolerance is necessary for crops to thrive in temperate regions where cold 3 snaps and lower baseline temperatures place limits on life processes; this is 4 particularly true for crops of tropical origin such as maize. Photosynthesis is often 5 adversely affected by chilling stress, yet the maintenance of photosynthesis is 6 essential for healthy growth and development, and most crucially for yield. In this 7 review we describe the physiological basis for enhancing chilling tolerance of 8 photosynthesis in maize by examining nine key responses to chilling stress. We 9 synthesise current knowledge of genetic variation for photosynthetic chilling 10 tolerance in maize with respect to each of these traits and summarise the extent to 11 which genetic mapping and candidate genes have been used to understand the 12 genomic regions underpinning chilling tolerance. Finally, we provide perspectives on 13 the future of breeding for photosynthetic chilling tolerance in maize. We advocate 14 for holistic and high-throughput approaches to screen for chilling tolerance of 15 photosynthesis in research and breeding programmes in order to develop resilient 16 crops for the future. 17 18 Keywords 19 20 Breeding, Chilling Stress, Cold Stress, Chilling Tolerance, Cold Tolerance, Genetics, 21 Maize, Photosynthesis, Quantitative Trait Loci (QTL), Spectroscopy 22 3 Introduction 23 24 Temperature is a key determinant of plant species distribution (Osmond et al., 1987; 25 Nievola et al., 2017), and our planet is experiencing a rise in the frequency and 26 severity of extreme temperature events (IPCC, 2018). At the same time, the world’s 27 population is increasing rapidly, demanding a concomitant increase in global food 28 production which will depend in part upon improved photosynthesis (Ort et al., 29 2015; Simkin et al., 2019). Whilst populations are stable or decreasing in many 30 countries that grow maize, improving photosynthesis is nevertheless of relevance for 31 maintaining crop yields in the context of temperature stresses exacerbated by 32 climate change. In cereal crops, reproduction is the most temperature-sensitive 33 growth stage (Yoshida et al., 1981) making temperature stress a critical limitation on 34 yield and therefore of direct relevance for food production; plant establishment and 35 vegetative growth are also susceptible to temperature stress. Chilling temperature 36 stress in particular is a strong limiting factor on plant growth and survival in 37 temperate regions, where it is the primary stress impacting germination as well as 38 affecting subsequent growth and development including crop production (Revilla et 39 al., 2005; Sanghera et al., 2011). Chilling tolerance, including chilling tolerance of 40 photosynthesis, is therefore essential if plants are to survive and even to thrive in 41 such conditions. Improving our understanding of photosynthetic chilling tolerance in 42 crop plants is thus both critical and timely for maintaining and increasing food 43 production to support our growing global population. 44 45 4 Stress affects gene expression, metabolism, physiology and morphology (Krasensky 46 and Jonak, 2012). Chilling tolerance involves physiological or morphological 47 adaptations to combat chilling stress, in contrast to chilling avoidance which is 48 achieved by seed or vegetative dormancy (Revilla et al., 2005). Chilling tolerance can 49 occur at different timescales, which may be broadly arranged into three categories. 50 The longest of these is adaptation to chilling stress, which occurs when plants have 51 evolved to deal with perennially cold conditions; one example is evergreen trees 52 downregulating photosynthesis (Savitch et al., 2002). Next, in contrast to 53 evolutionary adaptation, acclimation to chilling stress occurs when plants are grown 54 under cold conditions that they do not necessarily always experience; chilling 55 tolerant species are those which are able to acclimate to cold temperatures 56 (Ensminger et al., 2006). This acclimation involves the employment of survival 57 strategies that are not constitutively expressed under all growth conditions, in 58 response to a chronic chilling stress that persists for much of the season. Finally, on 59 the shortest timescale, tolerance to acute chilling stress describes resilience to cold 60 snaps – short periods of unexpected or unseasonal cold weather to which plants 61 may not already be acclimated (Hüner et al., 2016). This review focuses on chilling 62 tolerance in maize (Zea mays L.), a species in which the chilling response has been 63 much studied in order to facilitate the growth of this important crop in temperate 64 regions. Maize is the most grown cereal crop in the world, making its temperature 65 response a critical aspect of global food security. Since maize is not adapted to deal 66 with low temperatures, we consider the responses of maize to chronic and acute 67 chilling stresses caused by cool seasons or cold snaps respectively. 68 69 5 Plant species which originated in tropical regions are often especially sensitive to 70 chilling stress (Sanghera et al., 2011), and maize is no exception (Miedema, 1982). 71 We define chilling stress as the presence of suboptimal cool temperatures above 0°C. 72 Chilling and freezing temperatures impose stress in different ways: chilling stress 73 imposes a direct temperature stress whilst freezing stress, which occurs at sub-zero 74 temperatures, causes osmotic stress via the dehydration of cells when extracellular 75 ice crystals are formed (Hincha and Zuther, 2020); the two stresses are both 76 genetically and physiologically distinct (Revilla et al., 2005). The specific 77 temperatures causing chilling stress vary between species, as well as between 78 different growth stages and different organs of the plant (Revilla et al., 2005). For 79 example, roots are especially sensitive to chilling stress and restrictions on root 80 growth can lead to downstream effects such as a reduced supply of water and 81 nutrients later in development. The seed imbibition and photosynthetic initiation 82 phases are the most chilling sensitive phases within the seed germination and 83 seedling growth period (Revilla et al., 2005). Chilling stress can occur at 84 temperatures ranging from 0 to 15°C (Revilla et al., 2005; Zhu et al., 2007; Sanghera 85 et al., 2011; Miura and Furumoto, 2013) and temperatures within this range are 86 used for experimental work imposing chilling stress on maize (Hu et al., 2017; 87 Frascaroli and Revilla, 2019). 88 89 Maize originated in the tropics but has been adapted to a range of climates. 90 European varieties of maize, such as varieties in the ‘Flint’ race, can display greater 91 chilling tolerance than those of tropical origin. Indeed, Flint lines are often used in 92 northern European maize breeding to provide chilling tolerance (Riva-Roveda et al., 93 6 2016). In temperate regions, where maize production has been increasing for several 94 decades (Fracheboud et al., 1999; Frascaroli and Revilla, 2019), early planting 95 increases plant biomass and reduces exposure to drought and parasites and the 96 associated canopy coverage decreases competition with weeds. However, early 97 planting also increases seedling stress from chilling and disease. Overall, maize 98 establishment is more difficult in temperate regions (Jompuk et al., 2005). 99 100 Chilling stress in maize is already considered to occur at temperatures below 10-101 15°C (Hu et al., 2017; Frascaroli and Revilla, 2019). Generally speaking, temperatures 102 below 15°C slow growth, with this threshold increasing to 20°C in more established 103 plants, while temperatures below 5°C cause further cell and tissue damage, and 104 injury to seeds and seedlings (Frascaroli and Revilla, 2019). Temperatures below 105 10°C badly affect maize germination (Janowiak et al., 2002) and photosynthesis 106 (Foyer et al., 2002), although it should be noted that maize varieties display 107 significant variation in chilling tolerance as discussed below. In an agricultural setting, 108 severe chilling stress can occur below 8°C and maize should therefore ideally be 109 sown when temperatures exceed this threshold (Sobkowiak et al., 2014, 2016; 110 Jończyk et al., 2017). 111 112 Chilling stress has multiple effects on plant morphology and function (Fig. 1). Chilling 113 reduces plant establishment, growth and reproduction, and leads to wilting, 114 chlorosis and necrosis (Revilla et al., 2005). Chlorosis can be linked to cell membrane 115 disruption (Miedema, 1982); properties of the cell wall and membrane are important 116 for chilling tolerance (Sanghera et al., 2011; Frascaroli and Revilla, 2019). Chilling 117 7 stress affects metabolism, root growth and morphology, leaf area, number of days to 118 emergence, germination rate, chlorophyll and the efficiency of photosystem II, ΦPSII 119 (Janowiak et al., 2002; Hund et al., 2007; Frascaroli and Revilla, 2019). Cell 120 membrane disruption, chlorophyll bleaching and decreased ΦPSII contribute to 121 lowered rates of photosynthesis, impacting growth and productivity. 122 123 Chilling tolerance is a complex polygenic trait (Tokuhisa and Browse, 1999; 124 Thomashow, 2001) and its genetic regulation is not well understood (Frascaroli and 125 Revilla, 2019). Chilling tolerance in maize is regulated independently at different 126 growth stages (Hodges et al., 1997; Revilla et al., 2000) and furthermore there are 127 interactions between genotype and environment (Fracheboud et al., 2004; Revilla et 128 al., 2005; Presterl et al., 2007). Genes involved in chilling tolerance may be identified 129 through transcriptomic, proteomic or genomic approaches (Frascaroli and Revilla, 130 2019). Variation in traits of interest may be mapped to the genome using genomic 131 markers such as single nucleotide polymorphisms (SNPs) or quantitative trait loci 132 (QTL). SNPs are particularly useful for performing genome wide association studies 133 (GWAS) which can identify QTL and increase the resolution of QTL mapping (Hu et al., 134 2017; Frascaroli and Revilla, 2019; Li et al., 2021). Genetic mapping performed using 135 SNPs can also be used for marker assisted selection and to identify candidate genes 136 (Miculan et al., 2021; Waqas et al., 2021). Once identified, candidate genes relating 137 to physiological traits of interest may be classified according to functional 138 characteristics using gene ontology (GO) terms; GO-term annotations are now 139 available for all of the protein-coding genes in maize (Wimalanathan et al., 2018). 140 141 8 In this review we examine the physiology of photosynthetic chilling tolerance in 142 maize, genetic variation in photosynthetic chilling tolerance, and the genetic 143 elements underpinning this variation, in order to address the question: Can we 144 improve the chilling tolerance of maize photosynthesis through breeding? 145 9 Physiology of photosynthetic chilling tolerance 146 147 In order to survive a period of chilling stress, plants must adjust their physiological 148 processes in order to minimise damage. Maize plants display a range of chilling 149 responses (Fig. 1), and these occur on different timescales. In this section, we 150 examine the major physiological responses to chilling stress in maize, including both 151 immediate and longer-term responses, which enable plants to react to acute and 152 chronic chilling stress. Some responses are indicative of protective mechanisms that 153 mitigate the effects of chilling stress, whilst other responses reveal that damage has 154 already occurred. We outline three categories of physiological response to chilling 155 stress, organised according to the timescales in which they have been reported to 156 occur: photosynthetic responses, photoprotective responses, and signalling and 157 developmental responses. Finally, to conclude this section we consider the most 158 appropriate physiological measurements for screening natural genetic variation in 159 photosynthetic chilling tolerance. 160 161 Photosynthetic responses to chilling – carbon assimilation 162 Maize carries out C4 photosynthesis, which involves a biochemical carbon-163 concentrating mechanism that helps to increase photosynthetic efficiency, especially 164 under hot and dry conditions. Atmospheric CO2 equilibrates with bicarbonate and is 165 firstly fixed – via phosphoenol-pyruvate carboxylase (PEPC; BRENDA:EC4.1.1.31) – 166 into the 4-carbon metabolite oxalo-acetate in the mesophyll. Oxalo-acetate is 167 converted to malate which diffuses along a concentration gradient inwards from the 168 mesophyll to the bundle sheath cells. In the chloroplasts of the bundle sheath cells 169 10 where Rubisco (BRENDA:EC4.1.1.39) is compartmentalised, decarboxylation of 170 malate mediated by NADP-ME (BRENDA:EC1.1.1.40) releases CO2 while reducing 171 NADP+ to NADPH. This carbon-concentrating mechanism augments the CO2:O2 ratio 172 and thus increases the efficacy of RuBP carboxylation by Rubisco in the Calvin-173 Benson-Bassham cycle by competitive inhibition of RuBP oxygenation. 174 175 The initial physiological responses to chilling stress in maize are related to carbon 176 assimilation. Firstly, the capacity and rate of net CO2 assimilation decrease (Fig. 2) 177 when plants are temporarily exposed to chilling stress. This can already be observed 178 after two hours’ exposure to 4°C chilling stress and was more pronounced after a 179 longer chilling stress of 16 hours (Ying et al., 2002), as well as being observed after a 180 chilling stress of six hours (Aguilera et al., 1999). In both of these studies, 181 measurements of CO2 assimilation were made during a recovery period following the 182 chilling stress period. The decrease in net CO2 assimilation rate is a highly sustained 183 response, which has been reported in many studies after one day (Dwyer and 184 Tollenaar, 1989; Ying et al., 2000; Aroca et al., 2001; Riva-Roveda et al., 2016), two 185 to three days (Ying et al., 2000) and eight days of chilling stress (Lee et al., 2002). The 186 measurements made by Dwyer and Tollenaar and by Ying et al. were carried out 187 during recovery after chilling stress whilst the other studies performed 188 measurements during the chilling stress treatment indicating that both 189 measurements during and after a chilling stress period may be used to measure 190 decreased CO2 assimilation that occurs during chilling and persists during recovery. A 191 decrease in CO2 assimilation rate was also reported in several studies which imposed 192 a chilling stress for the duration of the experimental period (Nie and Baker, 1991; 193 11 Kingston-Smith et al., 1999; Ying et al., 2002; Fracheboud et al., 2004; Zaidi et al., 194 2010; Rodríguez et al., 2014). 195 196 Various mechanisms may contribute to the sustained decrease in CO2 assimilation 197 including reduced enzyme activity, collapse of metabolic gradients between 198 mesophyll and bundle sheath cells, damage to the photosystems, and increased 199 diffusive limitations to CO2 uptake. Photosynthetic enzyme activities are often 200 reduced under chilling stress (Avila et al., 2018). The activities of FBPase 201 (BRENDA:EC3.1.3.11), Rubisco and PEPC decrease in chilled maize leaves (Kingston-202 Smith et al., 1997). At cooler temperatures, the speed of atomic movement and the 203 rate of collisions decrease; many enzymatic processes are attuned to operate best 204 within a range of optimal temperatures and will therefore perform relatively poorly 205 outside of the relevant range. Rubisco has been speculated to be especially limiting 206 in chilling conditions in C4 species, since C4 plants contain less Rubisco and because 207 Rubisco is operating closer to its maximum capacity due to the high concentration of 208 CO2 created by the carbon-concentrating mechanism (Sage and McKown, 2006). 209 Furthermore, enhanced degradation of photosynthetic gene products under chilling 210 stress reduces the amounts of enzymes in the leaf: protein breakdown is increased 211 at low temperatures, reviewed by Sales et al. (2021). Specifically, the photosynthetic 212 enzymes PPDK (BRENDA:EC2.7.9.1), PEPC, and Rubisco break down more easily 213 under chilling conditions in C4 species (Kingston-Smith et al., 1997; Du et al., 1999; 214 Chinthapalli et al., 2003). This increased lability means that greater enzyme synthesis 215 is required to maintain a given activity, and therefore decreases the overall enzyme 216 activity across the leaf. The amounts and activities of enzymes can also trade off 217 12 against one another as part of the chilling stress response. For example, a decrease 218 in Rubisco content coupled with an observed increase in Rubisco activation state 219 may indicate an upregulation of activation in order to compensate for the lower 220 enzyme content observed during a chronic chilling stress experiment in maize 221 (Kingston-Smith et al., 1999). 222 223 While stomatal conductance is usually not a strong constraint to photosynthesis in 224 maize, it also decreases strongly under chilling conditions (Lee et al., 2002), which 225 may enhance the diffusional limitation to CO2 uptake. However, since CO2 226 assimilation and stomatal conductance are strongly coordinated, the stomatal 227 closure response is more likely to be a reflection of the chilling-induced decreases in 228 CO2 assimilation. 229 230 Photosynthetic responses to chilling – electron transport 231 As well as a decrease in net CO2 assimilation, chilling stress causes a decrease in the 232 operating efficiency of PSII in the light (ΦPSII), which is derived from measurements 233 of chlorophyll fluorescence (Maxwell and Johnson, 2000; Baker, 2008). 234 Downregulation of ΦPSII in response to chilling occurs in parallel with changes in CO2 235 assimilation, being observed as early as two hours into chilling stress (Fig. 2), both 236 directly measured during chilling stress (Fracheboud et al., 2002) and via a decrease 237 in the maximum quantum efficiency of PSII photochemistry, Fv/Fm measured during 238 recovery following two hours of chilling (Ying et al., 2002). Decreases in Fv/Fm were 239 also reported at two, four and eight hours into a chilling stress period with greater 240 decreases observed as time progressed (Dolstra et al., 1994). The downregulation of 241 13 ΦPSII has also been reported a few hours after the imposition of chilling stress 242 (Sobkowiak et al., 2014); after one day of chilling stress (Sobkowiak et al., 2014, 243 2016); after two days (Janowiak et al., 2002; Sobkowiak et al., 2014; Urrutia et al., 244 2021); four days (Urrutia et al., 2021); five days (Sobkowiak et al., 2014); six days 245 (Urrutia et al., 2021); eight days (Lee et al., 2002); and ten days (Kościelniak et al., 246 2005). Each of these results was obtained during the period of chilling stress, 247 although the study by Janowiak et al. also included measurements made during a 248 recovery period which are not reported here. In the case of the measurements by 249 Kościelniak et al., the chilling stress was even augmented at the time of 250 measurement, with measurements made at 6°C following a period of ten days at 251 15°C. As has been demonstrated for CO2 assimilation rate, the chilling-induced 252 decrease in ΦPSII persists during prolonged periods of chilling stress, being reported 253 by studies imposing chilling stress for the duration of the experiment (Fracheboud et 254 al., 1999, 2004; Kingston-Smith et al., 1999; Hund et al., 2007). 255 256 Balancing ΦPSII with CO2 assimilation enables plants to maintain an appropriate 257 energy balance, regulated by redox and pH changes as well as calcium signalling 258 initiated by changes in plasma membrane fluidity (Ensminger et al., 2006). CO2 259 assimilation and ΦPSII are correlated, but the relationship between them is not 260 always constant. For example, the relationship between CO2 assimilation and ΦPSII 261 can change under chilling conditions, with higher values of ΦPSII relative to CO2 262 assimilation (Fryer et al., 1998). However, this is not always the case, with other 263 studies reporting a more sustained relationship between CO2 assimilation and ΦPSII 264 during chilling stress (Kingston-Smith et al., 1997; Foyer et al., 2002), particularly 265 14 when irradiance is stable (Earl and Tollenaar, 1998). PSII is chronologically the first of 266 two photosystems in linear photosynthetic electron transport, which produces ATP 267 and reductant (NADPH) for subsequent use in the C4 acid shuttle and the Calvin-268 Benson-Bassham cycle to assimilate CO2 into carbohydrates. PSII is typically thought 269 to be more susceptible to chilling stress than PSI (Kočová et al., 2009). The PSII 270 reaction centre protein D1 is easily damaged, leading to photoinhibition and reduced 271 rates of photosynthesis; this can occur in chilling conditions particularly when 272 irradiance is high (Farage and Long, 1987). However, PSI is also easily damaged 273 under chilling conditions and sharp fluctuations in light intensity (Kono et al., 2014). 274 275 Downregulation of photosynthetic electron transport may not just be a result of run-276 away damage to the photosystems under chilling conditions. Instead, reversible 277 downregulation of PSII activity via non-photochemical quenching (NPQ), or more 278 long-term via halting the D1 protein repair cycle may also be initiated under chilling 279 conditions in order to balance carbon sources and sinks and to reduce the 280 potentially damaging effects of excessive light energy and concomitant production of 281 reactive oxygen species (Ensminger et al., 2006). Chilling stress reduces the 282 metabolic sink, and photosynthesis must respond in order to maintain an 283 appropriate carbon source:sink balance which is essential for maintaining healthy 284 growth (Ensminger et al., 2006; Burnett et al., 2016; White et al., 2016). Evidence for 285 this hypothesis comes from other plant species including evergreens and Arabidopsis, 286 in which chilling acclimation led to alterations to the thylakoid membrane, sucrose 287 synthesis enzyme expression and Calvin cycle enzyme expression, all of which have 288 been identified as balancing regulators of the carbon source and sink enabling 289 15 photosynthetic acclimation to chilling stress (Hüner et al., 1998, 2003; Stitt and 290 Hurry, 2002). In maize, expression of a sucrose synthase increased in response to 291 chilling stress (Urrutia et al., 2021), as has been seen in Arabidopsis (Stitt and Hurry, 292 2002), and downregulation of the expression of photosynthetic enzymes is also 293 observed, as we outline in the following section. 294 295 Photosynthetic responses to chilling – gene expression 296 Following soon after the downregulation of CO2 assimilation and ΦPSII is a 297 downregulation of photosynthetic gene expression (Fig. 2). While not as rapid as the 298 decreases in CO2 assimilation and PSII operating efficiency, downregulation of 299 photosynthetic gene expression (i.e. the abundance of photosynthesis related 300 transcripts) has been reported as early as 4 hours after the beginning of chilling 301 stress (Li et al., 2019), after 12 hours in another study (Yu et al., 2021), and after one 302 day of chilling stress in several studies (Trzcinska-Danielewicz et al., 2009; Zhang et 303 al., 2009; Jończyk et al., 2017; Avila et al., 2018; Banović Đeri et al., 2021). A 304 decrease in photosynthetic protein accumulation occurs soon after, reported after 305 two days of chilling stress (Urrutia et al., 2021). This decrease may be caused by the 306 transcriptional or translational downregulation of photosynthetic genes leading to a 307 reduction in protein synthesis; by the damage and breakdown of photosynthetic 308 proteins discussed above; by damage to the cellular machinery responsible for the 309 synthesis and repair of proteins; or by a combination of these factors. Similarly to 310 the other photosynthetic responses detailed in this section, the downregulation of 311 the expression of genes involved in photosynthesis persists during longer periods of 312 chilling stress, being reported after six days (Szalai et al., 2018), seven days (Riva-313 16 Roveda et al., 2016), and in long term studies of chilling stress (Nie and Baker, 1991; 314 Kingston-Smith et al., 1999). This sustained response of downregulation of gene 315 expression contributes to the sustained low rates of photosynthesis observed over 316 long periods of chilling stress. 317 318 Photoprotective responses to chilling – NPQ, chlorophyll content and reactive oxygen 319 species 320 Since enzymatic reactions are more strongly affected than the photochemical 321 electron transfer processes on the thylakoid membrane, chilling can lead to over-322 reduction of electron transfer components, and can promote production of 323 damaging reactive oxygen species. As a result, exposure to chilling tends to induce 324 photoprotective responses to mitigate these issues. Three potentially 325 photoprotective responses can already be seen after one day of chilling stress in 326 chilling tolerant maize plants: increased levels of NPQ, a decrease in chlorophyll 327 content, and an alteration in antioxidant enzymes or oxidative damage (Fig. 2). 328 These responses should be interpreted with caution, as each of these potentially 329 photoprotective mechanisms may also be a reflection of damage caused by chilling 330 stress. 331 332 NPQ is a compound term that encompasses a range of different non-photochemical 333 quenching mechanisms to dissipate excitation energy in the light-harvesting 334 antennae (reviewed by Malnoë, 2018). Some forms of NPQ are readily reversible 335 such as energy-dependent quenching (qE), which is primarily controlled by the pH of 336 the thylakoid lumen. In contrast, photoinactivation of the PSII reaction centre 337 17 protein D1 gives rise to a sustained photoinhibitory qI-type quenching, i.e. quenching 338 which leads to a long-term depression of the quantum yield of CO2 fixation. Unlike qE 339 which may be activated or deactivated within minutes, qI is not rapidly reversible as 340 it requires molecular repair. A decrease in Fv/Fm after dark-acclimation indicates the 341 presence of photoinhibition (Fracheboud et al., 1999). Increases in NPQ have been 342 observed after one day of chilling stress in some maize lines (Fig. 2), and may 343 continue after an additional two or six days depending on the line (Riva-Roveda et al., 344 2016; measurements made during chilling stress). Further resolving the different 345 forms of NPQ that are specifically upregulated in response to chilling will be 346 important for elucidating the photoprotective or photoinhibitory nature of these 347 responses. 348 349 A strong decrease in leaf chlorophyll content can often be observed in young maize 350 plants grown under suboptimal temperature. This phenotype may be a 351 manifestation of excessive oxidative damage to chlorophylls in the light-harvesting 352 antennae leading to photobleaching, but may also form part of a reorganisation and 353 restructuring of the light harvesting capacity as a photoprotective response to 354 chilling stress (Ensminger et al., 2006). A decrease in chlorophyll content can already 355 be observed after one day of chilling stress (Avila et al., 2018). This effect has also 356 been reported after five days (Aroca et al., 2001); six days (Szalai et al., 2018); seven 357 days (Riva-Roveda et al., 2016) and after eight days (Lee et al., 2002; Fig. 2). These 358 measurements were all performed during the chilling stress period although Aroca 359 et al. (2001) also included a subsequent recovery period, not reported here. 360 18 Furthermore, a decrease in chlorophyll content is a highly sustained response to 361 chilling stress, with multiple studies reporting decreased chlorophyll content after a 362 chilling stress that was imposed for the whole life of the maize plants prior to 363 measurement, suggesting that the potential for acclimation may be limited (Nie and 364 Baker, 1991; Kingston-Smith et al., 1999; Fracheboud et al., 2004; Hund et al., 2007; 365 Rodríguez et al., 2008). 366 367 Closely intertwined with chilling effects on leaf chlorophyll content, alterations in 368 antioxidant capacity also manifest after one day of chilling stress (Fig. 2). Increased 369 antioxidant enzyme activities were found in a chilling tolerant maize variety (Aroca 370 et al., 2001), whereas the antioxidant molecule ascorbic acid decreased after 30 371 hours of chilling stress in chilling sensitive sweet-corn seedlings (Xiang et al., 2020), 372 both measured during chilling stress. Alterations to antioxidant capacity can be very 373 persistent in response to long-term chilling. Increases in several antioxidant enzyme 374 activities were observed across a range of maize genotypes in response to 26 days of 375 chilling stress. In this case, superoxide dismutase, ascorbate peroxidase, and 376 glutathione reductase all showed increased activity whilst the response of catalase 377 activity was dependent on the genotype (Kočová et al., 2009). These changes in anti-378 oxidant capacity may impact accumulation of reactive oxygen species. For example, 379 increased hydrogen peroxide levels were observed in leaves exposed to 14°C 380 (Kingston-Smith et al., 1999), which may reflect enhanced oxidative stress levels. In 381 maize, the localisation patterns of antioxidant enzymes between bundle sheath and 382 mesophyll tissue (Doulis et al., 1997) increase the propensity for oxidative damage 383 (Kingston-Smith et al., 1999; Foyer et al., 2002). Reduced metabolite transport 384 19 between the bundle sheath and mesophyll tissues under chilling conditions increases 385 oxidative stress in the bundle sheath, since antioxidant enzymes are primarily 386 localised in the mesophyll (Kingston-Smith and Foyer, 2000). 387 388 Signalling and developmental responses to chilling – ABA, leaf sugar content, leaf 389 expansion 390 Lastly, we outline three responses related to signalling and development that occur 391 in response to chilling stress (Fig. 2). The fastest of these three is an increase in the 392 level of abscisic acid (ABA) which was already observable after two days as well as 393 after four and five days of chilling stress, and correlates with chilling tolerance 394 (Capell and Dörffling, 1993; Janowiak and Dörffling, 1996; Janowiak et al., 2002) and 395 has also been confirmed under field conditions (Janowiak et al., 2003). It is well 396 known that ABA is involved in the response to drought stress and exhibits crosstalk 397 with several metabolic and regulatory pathways (Ensminger et al., 2006; 398 Sreenivasulu et al., 2012; Munemasa et al., 2015; Sah et al., 2016; Zhu, 2016). Guard 399 cells are subject to ABA regulation, which stimulates stomatal closure during drought. 400 In chilling stress conditions, ABA may therefore contribute to a sustained decrease in 401 stomatal conductance to CO2 such as has been reported by Lee et al. (2002). 402 403 While increased ABA levels can occur relatively rapidly, a longer-term response to 404 chilling stress can be seen in the leaf soluble sugar content, which has been reported 405 to increase after seven days of chilling, measured during the chilling stress period 406 (Riva-Roveda et al., 2016). This increase could be a result of a decrease in phloem 407 loading, due to chilling-induced restrictions on transport (Krapp and Stitt, 1995; 408 20 Ainsworth and Bush, 2011). Alternatively, the increase in foliar sugar content may be 409 a physiological response to maintain turgor pressure when water transfer from the 410 roots is impaired by chilling stress. The accumulation of foliar sugars initiates 411 negative feedback repression of photosynthesis (Krapp and Stitt, 1995; Smeekens et 412 al., 2010), which may contribute to the sustained reduction in net CO2 assimilation 413 discussed above. 414 415 Finally, long-term exposure to chilling leads to a pronounced reduction in growth 416 rate, which can be observed very clearly in a decline of leaf expansion rate. This 417 common phenotype is often included in studies examining plants over multiple days 418 of chilling (e.g. Riva-Roveda et al., 2016). The slowing of leaf expansion and 419 appearance rate can be striking. For example, the time taken to reach the leaf 8 420 stage (the growth stage at which leaf 8 is the most recent fully expanded leaf, where 421 leaf 8 is the eighth leaf to appear on the plant) was tripled after eight days of chilling 422 stress at 15/13°C at the leaf 7 stage compared to plants grown under control 423 conditions (Lee et al., 2002). To account for the decreased rate of leaf expansion 424 under long-term chilling conditions, many studies compare control and chilling-425 treated plants at the same developmental stage rather than at the same time point 426 (Fracheboud et al., 1999, 2002, 2004). However, this can give rise to extreme age 427 differences between treatment and control groups, where the chilling-treated plants 428 can sometimes take twice as long to reach the same developmental stage (Rodríguez 429 et al., 2008). Whilst increases in foliar ABA and soluble sugars have not yet been 430 demonstrated to be sustained effects, a decrease in leaf expansion rate is clearly a 431 persistent effect during chilling stress. 432 21 433 Screening for chilling stress responses 434 Considering the nine responses outlined in this section (Fig. 2), the four most studied 435 components of the physiological response to chilling stress are consistent between 436 studies focused on exploring effects of chilling on physiological processes and studies 437 focused on examining genetic variation in chilling tolerance. These four components 438 are the three “photosynthesis” parameters, and chlorophyll content. However, the 439 degree to which each parameter is used varies between physiology- and genetics-440 focused studies. Assessing the studies of the photosynthetic chilling response in 441 maize reviewed here, in physiological studies net CO2 assimilation rate is the most 442 frequently studied parameter, followed by photosynthetic gene expression, ΦPSII and 443 chlorophyll content. In contrast, in genetics-focused studies, this order is reversed, 444 with chlorophyll content being the most frequently studied parameter, followed by 445 ΦPSII, photosynthetic gene expression and net CO2 assimilation. In both types of 446 study, the remaining five responses (NPQ; antioxidant enzymes or antioxidant 447 damage; ABA; leaf sugar content; and leaf expansion) are used relatively less 448 frequently. 449 450 The different emphasis on each of the three photosynthesis parameters and 451 chlorophyll content between physiology- and genetics-focused studies reflect the 452 different priorities of the two types of study. For studies of genetic variation, 453 chlorophyll content and ΦPSII provide rapid, relatively high-throughput proxies for 454 chilling stress which are useful for screening large populations and carrying out 455 genetic mapping, whilst measurements of net CO2 assimilation rate are less high-456 22 throughput but provide more physiological detail and are therefore favoured by 457 studies focusing on the physiological responses of maize to chilling stress. Regarding 458 the proxies for photosynthetic chilling tolerance favoured by genetics-focused 459 studies, we note that chlorophyll fluorescence is a particularly valuable screening 460 tool (Fracheboud et al., 1999; Baker, 2008). Specifically, ΦPSII provides a useful 461 means of distinguishing between chilling tolerant and chilling susceptible lines, and 462 has been used in initial breeding attempts to enhance chilling tolerance (Fracheboud 463 et al., 1999). Fluorescence measurements are non-destructive, facilitating repeated 464 measurements during an experimental time course. Chlorophyll content may be 465 measured destructively using pigment analysis following extraction in solvent, but 466 may also be estimated non-destructively from transmittance at a few specific 467 wavelengths with a SPAD meter or more elaborate spectrometry (Avila et al., 2018), 468 both providing great rapidity and the ability to repeat measurements throughout a 469 time course compared to biochemical pigment analysis. A major advantage of 470 chlorophyll fluorescence over chlorophyll content is the versatility and available 471 diversity of fluorescence measurements. Depending on the instrument and protocol 472 used, a measurement of a few minutes may suffice to provide Fv/Fm, ΦPSII and NPQ. 473 474 However, since both ΦPSII and chlorophyll content may be decreased during stress 475 for protective reasons or due to photodamage, it is advantageous to include a 476 metabolic component such as net CO2 assimilation rate or leaf sugar content in 477 parallel to allow more robust conclusions about the nature of the chilling response 478 to be drawn. The timescale of the response is also relevant: short-term 479 downregulation of ΦPSII or initiation of NPQ could be a photoprotective response, 480 23 whilst long-term differences in ΦPSII between genotypes are more likely to indicate 481 variation in the capacity for sustained photosynthesis under chilling conditions and 482 may therefore reveal chilling tolerance or susceptibility. 483 24 Genetic variation in chilling tolerance 484 485 Having established the primary physiological responses to chilling stress in maize, we 486 now examine the evidence for genetic variation within maize germplasm across 487 these responses. Our focus is on naturally occurring genetic variation, which 488 provides a useful pool of resources for breeding plants with greater chilling tolerance 489 of photosynthesis (Faralli and Lawson, 2020). Evidence for genetic variation in 490 photosynthetic chilling tolerance can become apparent whenever lines with 491 contrasting chilling tolerance are studied. Studies containing two – or a few – lines 492 may be used to identify differentially expressed genes (DEGs) in response to chilling 493 between tolerant versus susceptible lines. In contrast, large populations with 494 sufficient phenotypic variation and tractable genotypic variation are needed for the 495 identification of quantitative trait loci (QTL) or single nucleotide polymorphisms 496 (SNPs) that significantly correlate with variation in chilling tolerance. 497 498 Reflecting on the nine physiological responses identified in the previous section, 499 several studies have looked at gene expression changes in conjunction with chilling 500 treatments in tolerant and susceptible maize lines, and candidate genes have been 501 identified for chilling-related variation in CO2 assimilation rate, ΦPSII, photosynthetic 502 gene expression, chlorophyll content, antioxidant capacity, leaf sugar content and 503 morphology related to leaf expansion (summarized in Table 1), but not for NPQ or 504 ABA. In addition, several studies have used chilling-related variation in CO2 505 assimilation rate, ΦPSII, photosynthetic gene expression, NPQ and chlorophyll across 506 mapping populations to identify QTL for each of these traits. SNPs significantly 507 25 correlated with variation in CO2 assimilation, ΦPSII, chlorophyll and morphology 508 related to leaf expansion have also been identified (Table 1). In contrast, we could 509 not find any studies where genetic mapping was used for variation in antioxidant 510 capacity, reactive oxygen species (ROS) accumulation and oxidative damage, ABA, or 511 leaf sugar content in response to chilling (Table 1). Genetics-focused studies of 512 photosynthetic chilling tolerance typically measure CO2 assimilation, ΦPSII, 513 photosynthetic gene expression, and chlorophyll content. Considering these four 514 traits, some general trends emerge in studies that have examined genetic variation 515 in two or more genotypes (Table 1). Overall, decreases in CO2 assimilation, ΦPSII and 516 chlorophyll content are generally less pronounced in chilling tolerant genotypes 517 compared to chilling sensitive genotypes, indicating that lower values of ΦPSII and 518 chlorophyll content may more likely reflect the result of photodamage rather than 519 photoprotection in chilling sensitive lines. 520 521 Studies measuring multiple traits across chilling tolerant and chilling sensitive 522 genotypes frequently report relationships between physiological traits of interest. 523 These relationships provide information about whether certain responses might 524 indicate photoprotection or photodamage. For example, CO2 assimilation rate, ΦPSII 525 and chlorophyll content were all much lower in a chilling sensitive line than in a 526 chilling tolerant line under prolonged chilling stress in a study by Fracheboud et al. 527 (2004), indicating that reductions in ΦPSII and chlorophyll content are more likely 528 related to photodamage rather than protection. Similarly, in a study of two 529 genotypes there was a greater decrease in ΦPSII under chilling stress in the chilling 530 sensitive line compared to the chilling tolerant line (Sobkowiak et al., 2014). Similar 531 26 relationships were found between CO2 assimilation rate, ΦPSII and chlorophyll 532 content and chilling tolerance across a diverse collection of inbred lines (Lee et al., 533 2002), again suggesting that decreased ΦPSII is related to photodamage rather than 534 being a photoprotective response. Additionally, an examination of 19 lines 535 characterised for high or low ΦPSII under chilling stress showed that the “high ΦPSII” 536 lines had high CO2 assimilation rate, ΦPSII and chlorophyll content under chilling 537 stress (Hund et al., 2005). Similarly, QTL linked to higher ΦPSII under chilling stress 538 derived from a mapping population originated from the chilling tolerant parent 539 (Jompuk et al., 2005). And in a phenotypic screen for the effects of long term chilling, 540 a “favourable” allele was linked to higher ΦPSII (Fracheboud et al., 2002). Taken 541 together, these results indicate strongly that rather than lowering ΦPSII for 542 photoprotection, the maintenance of ΦPSII is an important aspect of photosynthetic 543 resilience to chilling stress. Similarly, most of the “favourable” alleles at the QTL 544 linked to a relatively smaller decrease in chlorophyll content under chilling stress 545 were derived from the chilling tolerant parent (Jompuk et al., 2005). This indicates 546 that in addition to limiting chilling-induced decreases in ΦPSII, the maintenance of 547 chlorophyll is also advantageous during chilling stress, and suggests that the 548 observed reduction in chlorophyll content may largely reflect photodamage rather 549 than photoprotection. 550 551 Whilst the evidence provided by these studies supports the hypothesis that reduced 552 ΦPSII and chlorophyll content are linked to photodamage, repeated measurements 553 made during a prolonged chilling stress also reveal a protective response that may 554 occur as a result of priming. Fracheboud et al. (2002) showed that following an initial 555 27 decrease in ΦPSII in leaf 1 in response to chilling stress, which is likely a result of 556 photodamage, ΦPSII in leaf 3, that was subsequently developed under chilling stress, 557 was also decreased. In this case the downregulation of ΦPSII may indeed be part of 558 photoprotective acclimatory responses. Interestingly, priming at a cool temperature 559 prior to the imposition of a more severe chilling stress of 8°C led to a less 560 pronounced reduction in ΦPSII at 8°C, compared to plants that had been exposed 561 directly to the 8°C treatment with no priming (Sobkowiak et al., 2016). This priming 562 was more beneficial in the chilling tolerant line than the two chilling sensitive lines 563 used in the study where the sensitive lines always showed a greater reduction in 564 ΦPSII than the tolerant line. 565 566 Regarding the expression of photosynthetic genes, Li et al. (2019) examined 567 transcriptional changes in a chilling tolerant and a chilling sensitive maize line. They 568 found that the number of DEGs was much greater in the tolerant line during the first 569 24 hours of chilling stress, with 1665 DEGs after 4 hours and 3970 DEGs after 24 570 hours; in the sensitive line there were 547 DEGs after 4 hours and 1766 DEGs after 571 24 hours. This may indicate either a more wide-ranging, or a more rapid, response in 572 the tolerant line, although a more prolonged time course would be required to 573 confirm this. Photosynthesis-related genes showed a faster response to chilling 574 stress in the tolerant line, whilst genes related to the light harvesting complexes 575 decreased after 4 hours in both lines indicating an early photoprotective response. 576 Interestingly, genes related to ΦPSII were downregulated after 24 hours of chilling 577 stress in the chilling sensitive line only, which suggests that the tolerant line was not 578 dependent on a photoprotective downregulation of ΦPSII. Indeed, in the sensitive 579 28 line, a greater decrease in Fv/Fm coupled with an increase in Fo (the minimum 580 fluorescence value measured after dark adaptation) indicated that photoinhibition 581 and photodamage had occurred. 582 583 Many studies examining changes in chlorophyll content in response to chilling have 584 identified both QTL and candidate genes, whilst few studies have identified 585 candidate genes relating to the chilling-induced decrease in net CO2 assimilation rate 586 (Table 1). CO2 assimilation is a complex trait, relying upon the amount, activation 587 state and activity of a range of enzymes as well as the physiological status of the leaf, 588 such as the status of the photosystems involved in the light reactions, 589 plasmodesmatal conductivity to facilitate metabolite transfer, phloem loading rate, 590 and – although only to a certain extent in C4 species – stomatal aperture. By contrast, 591 chlorophyll content depends primarily on the synthesis and breakdown of 592 chlorophyll, although of course the efficacy of chlorophyll in photosynthesis further 593 depends upon its binding and coordination within the light harvesting complexes. 594 Because the regulation of chlorophyll content is less complex than the regulation of 595 photosynthesis, it may be more straightforward to use chlorophyll content for the 596 identification of candidate genes to enhance chilling tolerance, rather than using CO2 597 assimilation or ΦPSII. Candidate genes involved in the regulation of chlorophyll 598 content under chilling stress have been identified, with more studies reporting 599 candidate genes for chlorophyll than any of the other traits, with the exception of 600 gene expression changes under chilling stress (Table 1). In spite of the relative 601 paucity of candidate genes related to net CO2 assimilation or to ΦPSII under chilling 602 conditions, the fact that many studies have identified QTL (or SNPs) for these two 603 29 traits suggests that it will be possible to establish some candidate genes in the near 604 future. The relative contribution of these QTL to the level of each trait in response to 605 chilling and the persistence of this contribution in different genomic backgrounds 606 and across different environments will be important determinants of their utility in 607 breeding programs. 608 609 Although several of the photosynthetic and photoprotective responses to chilling 610 have already been used for genetic mapping studies, this is not the case for variation 611 in antioxidant capacity in response to chilling. Candidate genes involved in 612 antioxidant capacity were identified both as a result of mapping variation in chilling 613 tolerance indices (Huang et al., 2013) as well as by making transcriptomic 614 comparisons between tolerant and sensitive lines (Sobkowiak et al., 2014; Jończyk et 615 al., 2021), but their involvement awaits further experimental verification since 616 antioxidant capacity was not directly measured in any of these studies. Following up 617 this work with direct measurements of antioxidant capacity, or with genetic mapping 618 of variation in antioxidant capacity in response to chilling may provide another piece 619 of the puzzle as we move towards a more complete understanding of chilling 620 tolerant photosynthesis in maize. Interestingly, the accumulation of zeaxanthin was 621 negatively correlated with chilling tolerance in a study of maize genotypes differing 622 in chilling tolerance (Fracheboud et al., 2002). While the accumulation of zeaxanthin 623 is associated with a sustained form of NPQ (qZ; Nilkens et al., 2010), it is also a 624 potent ROS scavenger (Havaux et al., 2007), which leaves two possible explanations 625 for the observed negative relationship. On the one hand, the impact of zeaxanthin 626 on NPQ may depress maize photosynthetic efficiency in response to chilling as 627 30 suggested by Fryer et al. (1995). Alternatively, the increased accumulation of 628 zeaxanthin in sensitive genotypes could reflect a greater need for photoprotection in 629 these genotypes. The fact that lower ΦPSII and CO2 assimilation across the sensitive 630 genotypes in Fracheboud et al. (2002) also correlated strongly with proxies for larger 631 light harvesting antennae, which would increase excitation pressure per PSII reaction 632 centre, would seem most consistent with the second explanation. 633 634 Overall, many QTL relating to the physiological components of photosynthetic 635 chilling tolerance in maize have been identified, particularly with respect to CO2 636 assimilation, ΦPSII, and chlorophyll content. However, it is striking that relatively few 637 candidate genes have been identified when considering the broad range of studies 638 examined in this review (Table 1). This may be due to the fact that many traits are 639 polygenic, meaning that whilst QTL may be readily identified, pinpointing genes of 640 interest that are responsible for the traits in question is altogether more difficult. 641 For example, CO2 assimilation is an emergent property that depends upon a plethora 642 of physiological and molecular players, meaning that a wealth of genes underpins 643 this complex trait. Likewise, candidate genes for ΦPSII are relatively rare and no 644 candidate genes for NPQ have been identified (Table 1). Whilst transcriptomic 645 analysis of photosynthetic gene expression by definition identifies the expression of 646 photosynthesis-related genes, even chlorophyll content – which is a comparatively 647 simple trait related to chilling tolerance of photosynthesis – does not have many 648 associated candidate genes in the studies reviewed here, whilst for leaf sugar 649 content the candidate genes that have been identified are involved in phloem 650 loading rather than being more directly involved in sugar metabolism (Table 1). 651 31 Finally, some genes relating to antioxidant activity and to leaf expansion have been 652 identified, but none for ABA with respect to chilling tolerance (Table 1). It should 653 also be noted that QTL mapping is much easier than the definitive identification of 654 candidate genes, and since the draft genome of maize was published relatively 655 recently (Schnable et al., 2009), the possibility of identifying candidate genes is 656 rather new in maize compared to model species such as Arabidopsis. Furthermore, 657 many of the studies reviewed here focused on meeting breeding objectives, for 658 which QTL are instrumental but the identification of specific candidate genes is 659 generally not necessary. From a physiological perspective, elucidating the causal 660 sequence for a trait increases the possibility of successfully understanding the 661 underlying mechanism, so studies focused on physiological goals may be more likely 662 to pursue the identification of candidate genes rather than QTL. 663 664 Looking to the future, there exists significant diversity in ΦPSII between breeding 665 groups and populations (Strigens et al., 2013) and this could be exploited for the 666 development of chilling tolerant germplasm. Future studies might investigate the 667 genetic basis of variation in the other physiological traits we have highlighted in this 668 review, and the contribution of this variation to chilling tolerance or susceptibility. 669 The identification of more candidate genes will also be important, as outlined above. 670 Due to the complexity of several responses with respect to photoprotection and 671 damage, the use of experimental time courses in combination with phenotyping 672 across the broader spectrum of physiological responses to chilling as outlined here 673 will be critical for appropriate interpretation and may lead to the identification of 674 more stable QTL and candidate genes. 675 32 High-throughput breeding approaches 676 677 Having examined the physiological basis for photosynthetic chilling tolerance and 678 the genetic variation for this tolerance revealed in a range of populations and 679 responses, we now return to our central question: Can we improve the chilling 680 tolerance of maize photosynthesis through breeding? Whereas most of the 681 responses to chilling appear to show intra-specific genetic variation in maize, 682 appropriate interpretation of this variation requires determination of several 683 responses in parallel across large populations. 684 685 Physiological breeding for improving photosynthetic chilling tolerance 686 Physiological breeding aims to incorporate physiological trait measurements into 687 breeding programmes (Reynolds and Langridge, 2016). Such measurements can be 688 more time-consuming and labour-intensive, but are valuable for understanding the 689 physiological responses of plants to different stresses, especially when combined 690 with powerful QTL analysis in the breeding context. High-throughput approaches for 691 measuring physiological traits are therefore of great benefit; two such approaches 692 are chlorophyll fluorescence, which has been discussed above, and reflectance 693 spectroscopy. While measurements of ΦPSII using chlorophyll fluorescence may be 694 readily applied in a high-throughput manner (Hund et al., 2005) and can be tailored 695 to specific traits of interest (Maxwell and Johnson, 2000; Baker, 2008; Murchie and 696 Lawson, 2013), several additional techniques to cover more of the nine key 697 responses to chilling in parallel are now available. In particular, reflectance 698 spectroscopy offers another high-throughput approach. A major advantage of this 699 33 technique is that similar to fluorescence techniques, a rapid measurement (~ 1s) 700 enables the simultaneous estimation of a suite of metabolic and physiological 701 parameters of interest via correlative models (Yendrek et al., 2017; Ely et al., 2019; 702 Burnett et al., 2021c,a,b). For example, following the development of training 703 datasets and models which are appropriate for the genotypes and traits of interest, 704 the maximum carboxylation rate of Rubisco (Serbin et al., 2012; Meacham-Hensold 705 et al., 2020), leaf protein and sugar content (Ely et al., 2019), ABA (Burnett et al., 706 2021b), and chlorophyll content (Yendrek et al., 2017) may all be predicted from a 707 single hyperspectral measurement. Taken together, these parameters provide a 708 more holistic picture of the physiological response to chilling stress and would 709 enable quantification of photoprotective mechanisms as well as foliar damage 710 caused by chilling. Chilling tolerance can trade off against other useful desired traits 711 in maize (Frascaroli and Revilla, 2019); this furthers the requirement for a holistic 712 perspective when breeding for chilling tolerance. 713 714 Hyperspectral reflectance measurements are rapid and, once equipment has been 715 purchased, the costs per measurement are negligible. Many options are available 716 including leaf clips for leaf-level measurements and unmanned aerial vehicle (UAV) 717 platforms for screening fields at the plot level. Currently, hyperspectral 718 measurements typically need calibration within each system of interest before they 719 may be used for trait identification. However, it is possible to predict the structural 720 trait leaf mass per unit area (LMA) using reflectance data alone (Serbin et al., 2019) 721 and in the future it will become increasingly feasible to predict traits of interest 722 based on generalised models once models have been trained on wider-ranging 723 34 datasets and the leaf structural and optical properties have been accounted for. This 724 will significantly augment the utility of hyperspectral reflectance for breeding 725 programmes. 726 727 A physiological breeding approach will be instrumental when dealing with multiple 728 complex stresses. Rarely does a single stress occur. Rather, the dynamic field 729 environment can impose stresses in combination, such as heat and drought stress 730 during hot summers, or chilling and high light stress in temperate spring seasons; 731 considering biotic stresses such as pathogens adds a further dimension. Interestingly, 732 plant responses to stresses often overlap or compound each other. For example, a 733 population of 233 maize RILs derived from a drought tolerant and drought sensitive 734 parent was subsequently shown to contain a large degree of segregation in chilling 735 tolerance, demonstrating strong overlap between chilling and drought stress 736 tolerance (Fracheboud et al., 2002). Levels of ABA and proline, which are involved in 737 responses to and alleviation of drought stress, have also been shown to be involved 738 in acclimation to chilling stress in maize (Dory et al., 1990; Xin and Li, 1993; Revilla et 739 al., 2005). Chilling temperature stress generates a distinct metabolic and molecular 740 fingerprint, but also leads to responses that are shared with other stresses (Geange 741 et al., 2021). Understanding the hallmark signs of enhanced tolerance to a 742 combination of stresses is essential for breeding maize for an increasingly chaotic 743 and unpredictable future climate. 744 745 746 747 35 Breeding for enhanced chilling tolerance must consider crop phenology and target 748 environment 749 The goal of a breeding programme must be carefully considered when designing 750 experiments destined to inform the selection and development of maize germplasm. 751 Both field and controlled environments have limitations when it comes to 752 conducting chilling stress experiments; combining both approaches, with multiple 753 years and locations, is recommended for understanding and exploiting the true 754 variation in maize chilling tolerance (Frascaroli and Revilla, 2019). The timing of the 755 chilling stress is also important. Breeding chilling tolerant maize able to withstand 756 long-term chilling temperatures and acclimate to chilling conditions may give a 757 different outcome than breeding maize able to withstand short-term ‘cold snaps’ in 758 otherwise mild conditions. Cold snaps at any stage of growth can impact yield – by 759 reducing germination, slowing vegetative growth and development, or inhibiting 760 reproductive processes. Chilling tolerance does not always increase yield and indeed 761 there can be a trade-off between yield and stress tolerance (Revilla et al., 2005) 762 although historic maize yield improvement has been shown to be strongly related to 763 enhanced stress tolerance (Tollenaar and Wu, 1999). Successful breeding for chilling 764 tolerance must consider which growth stage is of particular interest and determine 765 which trait or combination of traits to target. Improvements in resource use 766 efficiency are often only revealed when plants are in stressful conditions (Tollenaar 767 and Wu, 1999). In this context we note that chilling stress at the reproductive stage 768 in maize is relatively under-studied; and may be an important area for further 769 research in an increasingly erratic climate. 770 771 36 Transgenic approaches for improving chilling tolerance of photosynthesis 772 While this review focuses on pre-existing variation in chilling tolerance of 773 photosynthesis in maize, and the genomic regions related to this tolerance which 774 may be utilised in breeding programmes, it is worth noting that genetic modification 775 approaches also offer valuable tools for improving photosynthesis and chilling 776 tolerance. For example, increasing Rubisco and electron transport capacity can 777 improve the photosynthetic performance of C4 plants; Rubisco is predicted to have a 778 greater effect on chilling recovery than other photosynthetic enzymes in the C4 779 pathway (Sales et al., 2021). The overexpression of Rubisco large and small subunits, 780 in concert with Rubisco Assembly Factor 1 (RAF1), increased maize Rubisco content 781 by over 30% (although Rubisco activase is likely a vital factor for translating this 782 increased enzyme content into a proportional increase in photosynthetic activity); 783 this overexpression of Rubisco can speed recovery following chilling stress (Salesse-784 Smith et al., 2018). Transgenic introduction of chilling tolerant PPDK into maize 785 lowered the threshold for chilling stress in the extracted enzyme and increased 786 photosynthesis by 23% under chilling conditions of 8°C (Ohta et al., 2004, 2006) 787 whilst introducing the osmoprotectant molecule glycinebetaine transgenically into 788 maize increased photosynthesis and reduced chilling damage (Quan et al., 2004). 789 790 Transgenic work carried out in other species demonstrates useful proofs of concept, 791 although we acknowledge that a detailed discussion of this topic is outside the scope 792 of the present review. For example, the AlSAP gene from the grass Aeluropus 793 littoralis has been successfully expressed in rice where it increased photosynthesis 794 and stress tolerance when plants were exposed to a chilling treatment as well as 795 37 other abiotic stresses (Ben Saad et al., 2012). Work in Arabidopsis has shown that 796 the CBF/DREB1 transcription factors are important for the chilling response (Miura 797 and Furumoto, 2013), and transgenic CBF/DREB1 transcription factors from 798 Arabidopsis have been used to improve chilling tolerance in tobacco and wheat 799 (Sanghera et al., 2011). Multiple genes, including genes from the CBF/DREB1 family, 800 have been transgenically introduced into rice to increase chilling tolerance, 801 highlighting the complex nature of chilling tolerance and its regulation (da Cruz et al., 802 2013). 803 804 Finally, the activation of latent genes already present within the genome, and a 805 greater understanding of genetic regulatory mechanisms, are important elements of 806 increasing chilling tolerance (Revilla et al., 2005). Transgenic approaches may also be 807 used to investigate the presence and function of genes that already exist within the 808 species of interest. For example, a study overexpressing a stress-responsive binding 809 factor from the Antarctic grass Deschampsia antarctica in rice used RNA-seq to 810 identify a candidate set of genes involved in the rice chilling stress response, 811 putatively regulated by the D. antarctica binding factor (Byun et al., 2018). Finally, 812 gene editing using CRISPR/Cas9 can be used to introduce specific beneficial alleles 813 into germplasm (Waqas et al., 2021). 814 815 Expanding allelic diversity for chilling tolerance 816 Considering conventional breeding methods, broad genetic diversity is important for 817 breeding (Revilla et al., 2005), and this includes diversity encompassing pre-existing 818 variation in photosynthesis (Faralli and Lawson, 2020). Introducing germplasm from 819 38 varieties or wild crop relatives adapted to high altitude and/or low temperature 820 areas can aid chilling tolerance of crops (Sanghera and Wani, 2008). In maize, the use 821 of germplasm from different environments of origin is a useful means of increasing 822 allelic diversity for improving chilling tolerance. For example, it was shown that many 823 Mexican highland maize landraces contain several introgressions obtained from a 824 highland subspecies of the wild relative teosinte (Zea mays ssp. mexicana). One of 825 these introgressions, a large chromosome inversion segment, could indeed be linked 826 to increased chilling tolerance and improved photosynthesis under chilling 827 conditions, including increased ΦPSII and increased chlorophyll gene expression 828 (Crow et al., 2020). The use of maize lines developed in temperate regions may also 829 improve chilling tolerance. In a study comparing 598 European inbred lines, several 830 “favourable” alleles for ΦPSII were identified, especially across the European flint 831 lines (Revilla et al., 2016). Local landraces may be used to introduce additional 832 diversity into elite germplasm, but due to their heterozygous nature these are more 833 difficult to use directly for breeding. Recent efforts to create doubled-haploid lines 834 produced from landraces therefore provide a useful resource for understanding and 835 exploiting the genetic and phenotypic diversity available in maize landraces (Hölker 836 et al., 2019). 837 838 Finally, it will be important to integrate agronomic and genetic approaches to 839 achieve future food security (McKersie, 2015). Besides breeding for increased 840 resilience, agronomic techniques can be employed to increase chilling tolerance. For 841 example, the application of “climate-smart agriculture” regimes such as altered 842 planting times, the application of exogenous plant growth regulators, and seed 843 39 coating and seed priming can further help to mitigate the effects of low 844 temperatures (Waqas et al., 2021). Just as priming with a moderate chilling stress 845 can alleviate a severe temperature stress in maize plants (Capell and Dörffling, 1993; 846 Sobkowiak et al., 2016), seed priming has been shown to improve antioxidant levels 847 and growth under chilling stress (Li et al., 2017). 848 40 Conclusion 849 850 Whilst the relationship between photosynthesis and yield is complex, photosynthesis 851 is a major contributing factor to yield (Sarquís et al., 1998; Simkin et al., 2019) and 852 the chilling tolerance of photosynthesis is an important component of improved 853 performance of maize under chilling temperatures (Dwyer and Tollenaar, 1989; 854 Tollenaar and Wu, 1999). Here we have identified nine traits that are pivotal in the 855 maize chilling response: carbon assimilation; electron transport; the expression of 856 photosynthetic genes; non-photochemical quenching; chlorophyll content; reactive 857 oxygen species; abscisic acid (ABA); leaf sugar content; and leaf expansion. Since the 858 chilling tolerance of photosynthesis is a complex breeding goal with multiple 859 phenotypic and genotypic components, we advocate for a multi-trait holistic 860 approach that takes specific phonological and geographical considerations into 861 account for successful breeding for chilling tolerance of photosynthesis. Breeding for 862 increased chilling tolerance of photosynthesis by exploiting the substantial natural 863 genetic variation for traits aligned with key chilling responses will improve maize 864 yields in cooler climes and contribute to meeting the significant global food security 865 challenges faced by humankind. 866 867 Acknowledgements 868 869 A.C.B. and J.K. were supported by the UKRI-FLF (Future Leaders Fellowship) 870 MR/T042737/1 awarded to J.K. 871 872 41 Author Contributions 873 874 A.C.B. and J.K. conceptualized the manuscript; A.C.B. wrote the manuscript with 875 input from J.K.; A.C.B. and J.K. edited the manuscript and approved the final version. 876 42 References Aguilera C, Stirling C, Long S. 1999. Genotypic variation within Zea mays for susceptibility to and rate of recovery from chill-induced photoinhibition of photosynthesis. Physiologia Plantarum 106, 429–436. Ainsworth EA, Bush DR. 2011. Carbohydrate export from the leaf: a highly regulated process and target to enhance photosynthesis and productivity. Plant Physiology 155, 64–9. Allam M, Revilla P, Djemel A, Tracy WF, Ordás B. 2016. Identification of QTLs involved in cold tolerance in sweet × field corn. Euphytica 208, 353–365. Aroca R, Irigoyen JJ, Sánchez-Díaz M. 2001. Photosynthetic characteristics and protective mechanisms against oxidative stress during chilling and subsequent recovery in two maize varieties differing in chilling sensitivity. Plant Science 161, 719–726. Avila LM, Obeidat W, Earl H, Niu X, Hargreaves W, Lukens L. 2018. Shared and genetically distinct Zea mays transcriptome responses to ongoing and past low temperature exposure. BMC Genomics 19, 1–18. Baker NR. 2008. Chlorophyll fluorescence: a probe of photosynthesis in vivo. Annual Review of Plant Biology 59, 89–113. Bano S, Aslam M, Saleem M, Basra SMA, Aziz K. 2015. Evaluation of maize accessions under low temperature stress at early growth stages. Journal of Animal and Plant Sciences 25, 392–400. Banović Đeri B, Božić M, Dudić D, Vićić I, Milivojević M, Ignjatović-Micić D, Samardžić J, Vančetović J, Delić N, Nikolić A. 2021. Leaf transcriptome analysis of Lancaster versus other heterotic groups’ maize inbred lines revealed different regulation of cold-responsive genes. Journal of Agronomy and Crop Science, 1–13. Ben Saad R, Fabre D, Mieulet D, Meynard D, Dingkuhn M, Al-Doss A, Guiderdoni E, Hassairi A. 2012. Expression of the Aeluropus littoralis AlSAP gene in rice confers broad tolerance to abiotic stresses through maintenance of photosynthesis. Plant, Cell and Environment 35, 626–643. Bilska-Kos A, Grzybowski M, Jończyk M, Sowiński P. 2016. In situ localization and changes in the expression level of transcripts related to intercellular transport and phloem loading in leaves of maize (Zea mays L.) treated with low temperature. Acta Physiologiae Plantarum 38. Burnett AC, Anderson J, Davidson KJ, Ely KS, Lamour J, Li Q, Morrison BD, Yang D, Rogers A, Serbin SP. 2021a. A best-practice guide to predicting plant traits from leaf- level hyperspectral data using partial least squares regression. Journal of Experimental Botany 72, 6175–6189. Burnett AC, Rogers A, Rees M, Osborne CP. 2016. Carbon source–sink limitations differ between two species with contrasting growth strategies. Plant, Cell and Environment 39, 2460–2472. Burnett AC, Serbin SP, Davidson KJ, Ely KS, Rogers A. 2021b. Detection of the metabolic response to drought stress using hyperspectral reflectance. Journal of Experimental Botany 72, 6474–6489. Burnett AC, Serbin SP, Rogers A. 2021c. Source:sink imbalance detected with leaf- and canopy-level spectroscopy in a field-grown crop. Plant Cell and Environment 44, 43 2466–2479. Byun MY, Cui LH, Lee J, Park H, Lee A, Kim WT, Lee H. 2018. Identification of rice genes associated with enhanced cold tolerance by comparative transcriptome analysis with two transgenic rice plants overexpressing DaCBF4 or DaCBF7, isolated from Antarctic flowering plant Deschampsia antarctica. Frontiers in Plant Science 9, doi: 10.3389/fpls.2018.00601. Capell B, Dörffling K. 1993. Genotype-specific differences in chilling tolerance of maize in relation to chilling-induced changes in water status and abscisic acid accumulation. Physiologia Plantarum 88, 638–646. Chinthapalli B, Murmu J, Raghavendra A. 2003. Dramatic differences in the responses of phosphoenolpyruvate carboxylase to temperature in leaves of C3 and C4 plants. Journal of Experimental Botany 54, 707–714. Crow T, Ta J, Nojoomi S, Aguilar-Rangel MR, Rodríguez JVT, Gates D, Rellán-Álvarez R, Sawers R, Runcie D. 2020. Gene regulatory effects of a large chromosomal inversion in highland maize. PLoS Genetics 16, 1–28. da Cruz RP, Sperotto RA, Cargnelutti D, Adamski JM, de FreitasTerra T, Fett JP. 2013. Avoiding damage and achieving cold tolerance in rice plants. Food and Energy Security 2, 96–119. Dolstra O, Haalstra SR, van der Putten PEL, Schapendonk AHCM. 1994. Genetic variation for resistance to low-temperature photoinhibition of photosynthesis in maize (Zea mays L.). Euphytica 80, 85–93. Dory I, Böddi B, Kissimon J, Paldi E. 1990. Cold stress responses of inbred maize lines with various degrees of cold tolerance. Acta Agronomica Hungaria 39, 309–318. Doulis AG, Debian N, Kingston-Smith AH, Foyer CH. 1997. Differential localization of antioxidants in maize leaves. Plant Physiology 114, 1031–1037. Du Y-C, Nose A, Wasano K. 1999. Effects of chilling temperature on photosynthetic rates, photosynthetic enzyme activities and metabolite levels in leaves of three sugarcane species. Plant, Cell & Environment 22, 317–324. Dwyer L, Tollenaar M. 1989. Genetic improvement in photosynthetic response of hybrid maize cultivars, 1959-1988. Candian Journal of Plant Science 69, 81–91. Earl HJ, Tollenaar M. 1998. Relationship between thylakoid electron transport and photosynthetic CO2 uptake in leaves of three maize (Zea mays L.) hybrids. Photosynthesis Research 58, 245–257. Ely KS, Burnett AC, Lieberman-Cribbin W, Serbin SP, Rogers A. 2019. Spectroscopy can predict key leaf traits associated with source-sink balance and carbon-nitrogen status. Journal of Experimental Botany 70, 1789–1799. Ensminger I, Busch F, Hüner NPA. 2006. Photostasis and cold acclimation: Sensing low temperature through photosynthesis. Physiologia Plantarum 126, 28–44. Farage P, Long S. 1987. Damage to maize photosynthesis in field during periods when chilling is combined with high photon fluxes. In: Biggens J, ed. Progress in Photosynthesis Research, Vol. IV. The Netherlands: Martinus Nijhoff Publishers, 139– 142. Faralli M, Lawson T. 2020. Natural genetic variation in photosynthesis: an untapped resource to increase crop yield potential? Plant Journal 101, 518–528. Foyer CH, Vanacker H, Gomez LD, Harbinson J. 2002. Regulation of photosynthesis and antioxidant metabolism in maize leaves at optimal and chilling temperatures: Review. Plant Physiology and Biochemistry 40, 659–668. 44 Fracheboud Y, Haldimann P, Leipner J, Stamp P. 1999. Chlorophyll fluorescence as a selection tool for cold tolerance of photosynthesis in maize (Zea mays L.). Journal of Experimental Botany 50, 1533–1540. Fracheboud Y, Jompuk C, Ribaut JM, Stamp P, Leipner J. 2004. Genetic analysis of cold-tolerance of photosynthesis in maize. Plant Molecular Biology 56, 241–253. Fracheboud Y, Ribaut JM, Vargas M, Messmer R, Stamp P. 2002. Identification of quantitative trait loci for cold-tolerance of photosynthesis in maize (Zea mays L.). Journal of Experimental Botany 53, 1967–1977. Frascaroli E, Revilla P. 2019. Genomics of cold tolerance in maize. In: Bennetzen J,, Flint-Garcia S, Hirsch C, Tuberosa R, eds. The Maize Genome. Springer Nature Switzerland, 287–304. Fryer MJ, Andrews JR, Oxborough K, Blowers DA, Baker NR. 1998. Relationship between CO2 assimilation, photosynthetic electron transport, and active O2 metabolism in leaves of maize in the field during periods of low temperature. Plant Physiology 116, 571–580. Fryer M, Oxborough K, Martin B, Ort DR, Baker N. 1995. Factors associated with depression of photosynthetic quantum efficiency in maize at low growth temperature. Plant Physiology 108. Geange SR, Arnold PA, Catling AA, et al. 2021. The thermal tolerance of photosynthetic tissues: a global systematic review and agenda for future research. New Phytologist 229, 2497–2513. Havaux M, Dall’Osto L, Bassi R. 2007. Zeaxanthin has enhanced antioxidant capacity with respect to all other xanthophylls in Arabidopsis leaves and functions independent of binding to PSII antennae. Plant Physiology 145, 1506–1520. Hincha DK, Zuther E. 2020. Plant cold acclimation and winter survival. In: Hincha DK, Zuther E, eds. Plant cold acclimation. New York: Springer Nature, 1–7. Hodges D, Andrews C, Johnson D, Hamilton R. 1997. Sensitivity of maize hybrids to chilling and their combining abilities at two developmental stages. Crop Science 37, 850–856. Hölker AC, Mayer M, Presterl T, Bolduan T, Bauer E, Ordas B, Brauner PC, Ouzunova M, Melchinger AE, Schön CC. 2019. European maize landraces made accessible for plant breeding and genome-based studies. Theoretical and Applied Genetics 132, 3333–3345. Hu G, Li Z, Lu Y, et al. 2017. Genome-wide association study identified multiple genetic loci on chilling resistance during germination in maize. Scientific Reports 7, 10840. Huang J, Zhang J, Li W, Hu W, Duan L, Feng Y, Qiu F, Yue B. 2013. Genome-wide association analysis of ten chilling tolerance indices at the germination and seedling stages in maize. Journal of Integrative Plant Biology 55, 735–744. Hund A, Fracheboud Y, Soldati A, Frascaroli E, Salvi S, Stamp P. 2004. QTL controlling root and shoot traits of maize seedlings under cold stress. Theoretical and Applied Genetics 109, 618–629. Hund A, Frascaroli E, Leipner J, Jompuk C, Stamp P, Fracheboud Y. 2005. Cold tolerance of the photosynthetic apparatus: Pleiotropic relationship between photosynthetic performance and specific leaf area of maize seedlings. Molecular Breeding 16, 321–331. Hund A, Richner W, Soldati A, Fracheboud Y, Stamp P. 2007. Root morphology and 45 photosynthetic performance of maize inbred lines at low temperature. European Journal of Agronomy 27, 52–61. Hüner NPA, Dahal K, Bode R, Kurepin L V., Ivanov AG. 2016. Photosynthetic acclimation, vernalization, crop productivity and ‘the grand design of photosynthesis’. Journal of Plant Physiology 203, 29–43. Hüner N, Öquist G, Melis A. 2003. Photostasis in plants, green algae and cyanobacteria: the role of light harvesting antenna complexes. In: Green B, Parson W, eds. Advances in Photosynthesis and Respiration Light Harvesting Antennas in Photosynthesis. Dordrecht: Kluwer Academic Publishers, 401–421. Hüner N, Öquist G, Sarhan F. 1998. Energy balance and acclimation to light and cold. Trends in Plant Science 3, 224–230. IPCC. 2018. Global warming of 1.5°C: an IPCC special report on the impacts of global warming of 1.5°C above pre-industrial levels and related global greenhouse gas emission pathways, in the context of strengthening the global response to the threat of climate change. Geneva, Switzerland. Janowiak F, Dörffling K. 1996. Chilling of maize seedlings: changes in water status and abscisic acid content in ten genotypes differing in chilling tolerance. Journal of Plant Physiology 147, 582–588. Janowiak F, Luck E, Dörffling K. 2003. Chilling tolerance of maize seedlings in the field during cold periods in spring is related to chilling-induced increase in abscisic acid level. Journal of Agronomy and Crop Science 189, 156–161. Janowiak F, Maas B, Dörffling K. 2002. Importance of abscisic acid for chilling tolerance of maize seedlings. Journal of Plant Physiology 159, 635–643. Jompuk C, Fracheboud Y, Stamp P, Leipner J. 2005. Mapping of quantitative trait loci associated with chilling tolerance in maize (Zea mays L.) seedlings grown under field conditions. Journal of Experimental Botany 56, 1153–1163. Jończyk M, Sobkowiak A, Trzcinska-Danielewicz J, Skoneczny M, Solecka D, Fronk J, Sowiński P. 2017. Global analysis of gene expression in maize leaves treated with low temperature. II. Combined effect of severe cold (8°C) and circadian rhythm. Plant Molecular Biology 95, 279–302. Jończyk M, Sobkowiak A, Trzcinska-Danielewicz J, Sowiński P. 2021. Chromatin- level differences elucidate potential determinants of contrasting levels of cold sensitivity in maize lines. Plant Molecular Biology Reporter 39, 335–350. Kingston-Smith AH, Foyer CH. 2000. Bundle sheath proteins are more sensitive to oxidative damage than those of the mesophyll in maize leaves exposed to paraquat or low temperatures. Journal of Experimental Botany 51, 123–130. Kingston-Smith AH, Harbinson J, Foyer CH. 1999. Acclimation of photosynthesis, H2O2 content and antioxidants in maize (Zea mays) grown at sub-optimal temperatures. Plant, Cell and Environment 22, 1071–1083. Kingston-Smith A, Harbinson J, Williams J, Foyer C. 1997. Effect of chilling on carbon assimilation, enzyme activation, and photosynthetic electron transport in the absence of photoinhibition in maize leaves. Plant Physiology 114, 1039–1046. Kočová M, Holá D, Wilhelmová N, Rothová O. 2009. The influence of low- temperature on the photochemical activity of chloroplasts and activity of antioxidant enzymes in maize leaves. Biologia Plantarum 53, 475–483. Kono M, Noguchi K, Terashima I. 2014. Roles of the cyclic electron flow around PSI (CEF-PSI) and O2-dependent alternative pathways in regulation of the photosynthetic 46 electron flow in short-term fluctuating light in Arabidopsis thaliana. Plant & Cell Physiology 55, 990–1004. Kościelniak J, Janowiak F, Kurczych Z. 2005. Increase in photosynthesis of maize hybrids (Zea mays L.) at suboptimal temperature (15°C) by selection of parental lines on the basis of chlorophyll a fluorescence measurements. Photosynthetica 43, 125– 134. Krapp A, Stitt M. 1995. An evaluation of direct and indirect mechanisms for the ‘sink-regulation’ of photosynthesis in spinach: Changes in gas exchange, carbohydrates, metabolites, enzyme activities and steady-state transcript levels after cold-girdling source leaves. Planta 195, 313–323. Krasensky J, Jonak C. 2012. Drought, salt, and temperature stress-induced metabolic rearrangements and regulatory networks. Journal of Experimental Botany 63, 1593– 1608. Lee EA, Staebler MA, Tollenaar M. 2002. Genetic variation in physiological discriminators for cold tolerance - Early autotrophic phase of maize development. Crop Science 42, 1919–1929. Li J, Li D, Espinosa CZ, et al. 2021. Genome-wide analyses reveal footprints of divergent selection and popping-related traits in CIMMYT’s maize inbred lines. Journal of Experimental Botany 72, 1307–1320. Li M, Sui N, Lin L, Yang Z, Zhang Y. 2019. Transcriptomic profiling revealed genes involved in response to cold stress in maize. Functional Plant Biology 46, 830–844. Li Z, Xu J, Gao Y, et al. 2017. The synergistic priming effect of exogenous salicylic acid and H2O2 on chilling tolerance enhancement during maize (Zea mays L.) seed germination. Frontiers in Plant Science 8, 1–14. Malnoë A. 2018. Photoinhibition or photoprotection of photosynthesis? Update on the (newly termed) sustained quenching component qH. Environmental and Experimental Botany 154, 123–133. Maxwell K, Johnson GN. 2000. Chlorophyll fluorescence – a practical guide. Journal of Experimental Botany 51, 659–68. McKersie B. 2015. Planning for food security in a changing climate. Journal of Experimental Botany 66, 3435–3450. Meacham-Hensold K, Fu P, Wu J, et al. 2020. Plot-level rapid screening for photosynthetic parameters using proximal hyperspectral imaging. Journal of Experimental Botany 71, 2312–2328. Miculan M, Nelissen H, Ben Hassen M, Marroni F, Inzé D, Mario Enrico P, Dell’Acqua M. 2021. A forward genetics approach integrating genome-wide association study and expression quantitative trait locus mapping to dissect leaf development in maize (Zea mays). Plant Journal 107, 1056–1071. Miedema P. 1982. The effects of low temperature on Zea mays. Advances in Agronomy 35, 93–128. Miura K, Furumoto T. 2013. Cold signaling and cold response in plants. International Journal of Molecular Sciences 14, 5312–5337. Munemasa S, Hauser F, Park J, Waadt R, Brandt B, Schroeder JI. 2015. Mechanisms of abscisic acid-mediated control of stomatal aperture. Current Opinion in Plant Biology 28, 154–162. Murchie EH, Lawson T. 2013. Chlorophyll fluorescence analysis: a guide to good practice and understanding some new applications. 64, 3983–3998. 47 Nie GY, Baker NR. 1991. Modifications to thylakoid composition during development of maize leaves at low growth temperatures. Plant Physiology 95, 184–191. Nievola C, Carvalho C, Carvalho V, Rodrigues E. 2017. Rapid responses of plants to temperature changes. Temperature 4, 371–405. Nilkens M, Kress E, Lambrev P, Miloslavina Y, Müller M, Holzwarth AR, Jahns P. 2010. Identification of a slowly inducible zeaxanthin-dependent component of non- photochemical quenching of chlorophyll fluorescence generated under steady-state conditions in Arabidopsis. Biochimica et Biophysica Acta - Bioenergetics 1797, 466– 475. Ohta S, Ishida Y, Usami S. 2004. Expression of cold-tolerant pyruvate, orthophosphate dikinase cDNA, and heterotetramer formation in transgenic maize plants. Transgenic Research 13, 475–485. Ohta S, Ishida Y, Usami S. 2006. High-level expression of cold-tolerant pyruvate, orthophosphate dikinase from a genomic clone with site-directed mutations in transgenic maize. Molecular Breeding 18, 29–38. Ort DR, Merchant SS, Alric J, et al. 2015. Redesigning photosynthesis to sustainably meet global food and bioenergy demand. Proceedings of the National Academy of Sciences 112, 8529–8536. Osmond CB, Austin M, Berry J, Billings W, Boyer J, Dacey J, Nobel P, Smith S, Winner W. 1987. Stress physiology and the distribution of plants. BioScience 37. Presterl T, Ouzunova M, Schmidt W, Möller E, Röber F, Knaak C, Ernst K, Westhoff P, Geiger H. 2007. Quantitative trait loci for early plant vigour of maize grown in chilly environments. Theoretical and Applied Genetics 114, 1059–1070. Quan R, Shang M, Zhang H, Zhao Y, Zhang J. 2004. Improved chilling tolerance by transformation with betA gene for the enhancement of glycinebetaine synthesis in maize. Plant Science 166, 141–149. Revilla P, Butrón A, Cartea M, Malvar R, Ordás A. 2005. Breeding for cold tolerance. In: Ashraf M, Harris PJC, eds. Abiotic stresses. Plant resistance through breeding and molecular approaches. The Haworth Press, Inc., USA, 301–398. Revilla P, Malvar R, Cartea M, Butrón A, Ordás A. 2000. Inheritance of cold tolerance at emergence and during early season growth in maize. Crop Science 40, 1579–1585. Revilla P, Rodríguez VM, Ordás A, et al. 2016. Association mapping for cold tolerance in two large maize inbred panels. BMC Plant Biology 16, 1–10. Reynolds M, Langridge P. 2016. Physiological breeding. Current Opinion in Plant Biology 31, 162–171. Riva-Roveda L, Escale B, Giauffret C, Périlleux C. 2016. Maize plants can enter a standby mode to cope with chilling stress. BMC Plant Biology 16, 1–14. Rodríguez VM, Butrón A, Malvar RA, Ordás A, Revilla P. 2008. Quantitative trait loci for cold tolerance in the maize IBM population. International Journal of Plant Sciences 169, 551–556. Rodríguez VM, Butrón A, Rady MOA, Soengas P, Revilla P. 2014. Identification of quantitative trait loci involved in the response to cold stress in maize (Zea mays L.). Molecular Breeding 33, 363–371. Rodríguez VM, Velasco P, Garrido JL, Revilla P, Ordás A, Butrón A. 2013. Genetic regulation of cold-induced albinism in the maize inbred line A661. Journal of Experimental Botany 64, 3657–3667. 48 Sage RF, McKown AD. 2006. Is C4 photosynthesis less phenotypically plastic than C3 photosynthesis? Journal of Experimental Botany 57, 303–317. Sah SK, Reddy KR, Li J. 2016. Abscisic acid and abiotic stress tolerance in crop plants. Frontiers in Plant Science 7, 1–26. Sales CRG, Wang Y, Evers JB, Kromdijk J. 2021. Improving C4 photosynthesis to increase productivity under optimal and suboptimal conditions. Journal of Experimental Botany 72, 5942–5960. Salesse-Smith CE, Sharwood RE, Busch FA, Kromdijk J, Bardal V, Stern DB. 2018. Overexpression of Rubisco subunits with RAF1 increases Rubisco content in maize. Nature Plants 4, 802–810. Sanghera G, Wani S. 2008. Innovative approaches to enhance genetic potential of rice for higher productivity under temperate conditions of Kashmir. Journal of Plant Science Research 24, 99–113. Sanghera G, Wani S, Hussain W, Singh N. 2011. Engineering cold stress tolerance in crop plants. Current Genomics 12, 30–43. Sarquís JI, Gonzalez H, Sánchez De Jiménez E, Dunlap JR. 1998. Physiological traits associated with mass selection for improved yield in a maize population. Field Crops Research 56, 239–246. Savitch L, Leonardos E, Krol M, Jansson S, Grodzinski B, Hüner N, Öquist G. 2002. Two different strategies for light utilization in photosynthesis in relation to growth and cold acclimation. Plant Cell and Environment 25, 761–771. Schnable PS, Ware D, Fulton RS, et al. 2009. The B73 maize genome: complexity, diversity and dynamics. Science 326, 1112–1115. Serbin SP, Dillaway DN, Kruger EL, Townsend PA. 2012. Leaf optical properties reflect variation in photosynthetic metabolism and its sensitivity to temperature. Journal of Experimental Botany 63, 489–502. Serbin SP, Wu J, Ely KS, Kruger EL, Townsend PA, Meng R, Wolfe BT, Chlus A, Wang Z, Rogers A. 2019. From the Arctic to the tropics: multibiome prediction of leaf mass per area using leaf reflectance. New Phytologist 224, 1557–1568. Simkin AJ, López-Calcagno PE, Raines CA. 2019. Feeding the world: Improving photosynthetic efficiency for sustainable crop production. Journal of Experimental Botany 70, 1119–1140. Smeekens S, Ma J, Hanson J, Rolland F. 2010. Sugar signals and molecular networks controlling plant growth. Current Opinion in Plant Biology 13, 274–9. Sobkowiak A, Jończyk M, Adamczyk J, et al. 2016. Molecular foundations of chilling- tolerance of modern maize. BMC Genomics 17, 125. Sobkowiak A, Jończyk M, Jarochowska E, Biecek P, Trzcinska-Danielewicz J, Leipner J, Fronk J, Sowiński P. 2014. Genome-wide transcriptomic analysis of response to low temperature reveals candidate genes determining divergent cold-sensitivity of maize inbred lines. Plant Molecular Biology 85, 317–331. Sreenivasulu N, Harshavardhan VT, Govind G, Seiler C, Kohli A. 2012. Contrapuntal role of ABA: Does it mediate stress tolerance or plant growth retardation under long- term drought stress? Gene 506, 265–273. Stitt M, Hurry V. 2002. A plant for all seasons: alterations in photosynthetic carbon metabolism during cold acclimation in Arabidopsis. Current Opinion in Plant Biology 5, 199–206. Strigens A, Freitag NM, Gilbert X, Grieder C, Riedelsheimer C, Schrag TA, Messmer 49 R, Melchinger AE. 2013. Association mapping for chilling tolerance in elite flint and dent maize inbred lines evaluated in growth chamber and field experiments. Plant, Cell and Environment 36, 1871–1887. Szalai G, Majláth I, Pál M, Gondor OK, Rudnóy S, Oláh C, Vanková R, Kalapos B, Janda T. 2018. Janus-faced nature of light in the cold acclimation processes of maize. Frontiers in Plant Science 9, 1–17. Thomashow M. 2001. So what’s new in the field of plant cold acclimation? Lots! Plant Physiology 125, 89–93. Tokuhisa J, Browse J. 1999. Genetic engineering of plant chilling tolerance. In: Setlow JK, ed. Genetic Engineering 21. New York: Kluwer Academic/Plenum Publishers, 79–93. Tollenaar M, Wu J. 1999. Yield improvement in temperate maize is attributable to greater stress tolerance. Crop Science 39, 1597–1604. Trzcinska-Danielewicz J, Bilska A, Fronk J, Zielenkiewicz P, Jarochowska E, Roszczyk M, Jończyk M, Axentowicz E, Skoneczny M, Sowiński P. 2009. Global analysis of gene expression in maize leaves treated with low temperature. I. Moderate chilling (14°C). Plant Science 177, 648–658. Urrutia M, Blein-Nicolas M, Prigent S, et al. 2021. Maize metabolome and proteome responses to controlled cold stress partly mimic early-sowing effects in the field and differ from those of Arabidopsis. Plant Cell and Environment 44, 1504–1521. Waqas MA, Wang X, Zafar SA, Noor MA, Hussain HA, Azher Nawaz M, Farooq M. 2021. Thermal stresses in maize: Effects and management strategies. Plants 10, 1–23. White AC, Rogers A, Rees M, Osborne CP. 2016. How can we make plants grow faster? A source–sink perspective on growth rate. Journal of Experimental Botany 67, 31–45. Wimalanathan K, Friedberg I, Andorf CM, Lawrence-Dill CJ. 2018. Maize GO Annotation—Methods, Evaluation, and Review (maize-GAMER). Plant Direct 2, 1–15. Xiang N, Hu J, Wen T, Brennan MA, Brennan CS, Guo X. 2020. Effects of temperature stress on the accumulation of ascorbic acid and folates in sweet corn (Zea mays L.) seedlings. Journal of the Science of Food and Agriculture 100, 1694– 1701. Xin Z, Li P. 1993. Relationship between proline and abscisic acid in the induction of chilling tolerance in maize suspension-cultured cells. Plant Physiology 103, 607–613. Yendrek CR, Tomaz T, Montes CM, Cao Y, Morse AM, Brown PJ, McIntyre LM, Leakey ADB, Ainsworth EA. 2017. High-throughput phenotyping of maize leaf physiological and biochemical traits using hyperspectral reflectance. Plant Physiology 173, 614–626. Ying J, Lee EA, Tollenaar M. 2000. Response of maize leaf photosynthesis to low temperature during the grain-filling period. Field Crops Research 68, 87–96. Ying J, Lee EA, Tollenaar M. 2002. Response of leaf photosynthesis during the grain- filling period of maize to duration of cold exposure, acclimation, and incident PPFD. Crop Science 42, 1164–1172. Yoshida S, Satake T, Mackill D. 1981. High temperature stress in rice. IRRI Research Paper Series 67, 1–15. Yu T, Zhang J, Cao J, et al. 2021. Leaf transcriptomic response mediated by cold stress in two maize inbred lines with contrasting tolerance levels. Genomics 113, 782–794. 50 Zaidi P, Yadav M, Maniselvan P, Khan R, Shadakshari T, Singh R, Pal D. 2010. Morpho-physiological traits associated with cold stress tolerance in tropical maize (Zea mays L.). Maydica 55, 201–208. Zhang Y, Fu J, Gu R, Wang J, Chen X, Jia J, Zhang J, Wang G. 2009. Isolation and analysis of cold stress inducible genes in Zea mays by suppression subtractive hybridization and cDNA macroarray. Plant Molecular Biology Reporter 27, 38–49. Zhu JK. 2016. Abiotic stress signaling and responses in plants. Cell 167, 313–324. Zhu J, Dong C, Zhu J. 2007. Interplay between cold-responsive gene regulation, metabolism and RNA processing during plant cold acclimation. Current Opinion in Plant Biology 10, 290–295. 51 Table 1. Genetic mapping and candidate genes for nine physiological responses to chilling stress in maize. Synthesis of studies including more than one genotype and measuring physiological responses to chilling stress. Studies are grouped according to the order of responses presented in Fig. 2. When describing each study, only chilling temperatures are included; control temperatures are omitted for brevity. Studies are listed under each applicable category but only described at the first instance. Study Genetic variation Genetic mapping Candidate genes CO2 assimilation rate F2:3 population from chilling tolerant (ETH-DH7) and chilling sensitive (ETH-DL3) lines. 15/13°C for whole life following establishment; measured leaf 3. (Fracheboud et al., 2004) Yes Yes – QTL for carbon exchange rate (a measurement of CO2 assimilation)1 No 233 RILs from drought tolerant (Ac7643) and drought susceptible (Ac7729/TZSRW) lines. 15/13°C for whole life following establishment; measured leaf 3. (Fracheboud et al., 2002) Yes Yes – QTL for CO2 fixation; 8 regions with QTL for photosynthetic traits; pericentromeric region of chromosome 3 a key location2 No 226 F2:3 families from ETH-DH7 x ETH-DL3 and 168 F2:4 from Lo964 x Lo1016 (different chilling tolerance at germination and different root morphology). 15/13°C for 14 d following establishment; measured leaf 3. (Hund et al., 2005) Yes Yes – QTL for carbon exchange rate No 282 inbred lines. 8°C at germination. (Hu et al., 2017) Carbon exchange rate not measured directly Yes – SNPs related to carbon exchange rate in other studies Yes – identified 18 candidate genes in total3 49 inbred lines. 15/13°C at 7 leaf stage, measured leaf 8. (Lee et al., 2002) Yes No No Photosystem II operating efficiency (ϕPSII) F2:3 population from ETH-DH7 x ETH-DL3. Early and late sowing in the field provided chilling treatment. (Jompuk et al., 2005) Yes Yes – QTL for ϕPSII located on chromosomes 2, 4, 6, 8, 9 (most prominent on 6) No Population from chilling sensitive x tolerant inbred lines. 14/8°C for the Yes Yes – two QTL for maintenance of ϕPSII No 1 QTL for a range of traits explained between 37 and 54% of the phenotypic variance in this study. 2 QTL explained up to 20% of phenotypic variance in this study. 3 Of these 18 genes, 10 were supported by other studies and 3 were novel. 52 duration of the experiment. (Rodríguez et al., 2014) in chilling stress4 (Fracheboud et al., 2004) Yes Yes – QTL for ϕPSII No (Fracheboud et al., 2002) Yes Yes – QTL for ϕPSII No (Hund et al., 2005) Yes Yes – QTL for ϕPSII, located on different chromosomes in the different populations No 168 F2:4 families from Lo964 x Lo1016 (see above). 15/13°C for the duration of the experiment; measured at 1st leaf stage. (Hund et al., 2004) Yes Yes – 4 QTL for ϕPSII A locus for ϕPSII was identified 1 chilling tolerant and 1 chilling sensitive line. (ETH-DH7 and ETH-DL3). 8/6°C imposed for 14 h at 3rd leaf stage. (Sobkowiak et al., 2014) Yes Yes – DEGs adjacent to QTL for chlorophyll fluorescence Yes – overall, identified 66 genes responding differently between lines (DEGs) (Lee et al., 2002) Yes No No Two panels: 306 Dent lines and 292 Flint lines. 14/8°C for duration of experiment. (Revilla et al., 2016) Yes Yes – 2 SNPs for ϕPSII in chilling stress in Flint population (chromosomes 1, 4); QTL for ϕPSII Overall, more QTL for chilling tolerance were identified in the Flint panel Yes – performed GWAS and identified candidate genes 3 breeding groups, total 375 inbred lines. 16/13°C. (Strigens et al., 2013) Yes – significant differences in ϕPSII between the breeding groups Yes – identified 3 QTL for ϕPSII (2 under chilling stress, 1 only under optimal conditions) No Photosynthetic gene expression (Sobkowiak et al., 2014) Yes Yes – DEGs adjacent to QTL for C4 enzymes Yes (see above) 1 chilling tolerant (S68911) and 2 chilling sensitive lines (S160 and S50676). 14/12°C for 4 d then 8/6°C for 4 d at 3rd leaf stage. (Sobkowiak et al., 2016) Yes No Yes – GO enrichment identified photosynthetic genes 2 unrelated inbred lines: CG60, CG102. 14/2°C for 3 d at 2nd leaf stage; measured after 1 d chilling. (Avila et al., Yes No Yes – GO-term analysis identified photosynthetic genes downregulated in chilling stress 4 These two QTL explained 19% and 6% of phenotypic variance. 53 2018) 4 stress-sensitive “Lancaster” lines, 4 tolerant lines. 6/4°C for 24 h at 4th leaf stage. (Banović Đeri et al., 2021) Yes No Yes – 7 DEGs including photosynthetic genes. Differential expression between genotyoes and treatment/control and between genotypes 1 chilling tolerant (M54), 1 chilling sensitive (753F) line. 4°C chilling stress for up to 24 h at 4th leaf stage. (Li et al., 2019) Yes No Yes – chilling stress affected photosynthetic genes 1 chilling tolerant (B144), 1 chilling sensitive (Q319) line. 5°C chilling stress for 12 or 24 h at 3rd leaf stage. (Yu et al., 2021) Yes No Yes – upregulation of the D1 protein psb29 after 24 h (following initial decrease at 12 h) enabled B144 to protect PSII from photooxidation Non-photochemcial quenching (NPQ) (Fracheboud et al., 2002) Yes Yes – QTL for xanthophylls No A chilling sensitive inbred line (A661) and B73. 15°C for the duration of the experiment. (Rodríguez et al., 2013) Yes – lower xanthophylls in A661 No No Chlorophyll content 302 RILs from B73 x Mo17. 14/8°C for the experiment duration; measured after 30 d. (Rodríguez et al., 2008) Yes – measured chlorophyll using optical scale Yes – QTL identified on chromosomes 3 and 6, under chilling conditions only5 QTL on chromosome 6 may correspond to luteus11 locus (Fracheboud et al., 2004) Yes QTL identified on chromosome 3 No (Fracheboud et al., 2002) Yes Yes – QTL for chlorophyll No (Hund et al., 2005) Yes Yes – QTL for chlorophyll No (Hu et al., 2017) Chlorophyll not measured directly Yes – SNPs related to chlorophyll in other studies Yes – see above Two populations of field x sweet corn (B73 x P39, 179 RILs; B73 x IL14 h, 213 RILs). 14/10°C for the experiment duration. (Allam et al., 2016) Yes Yes – QTL for chlorophyll content No (Hund et al., 2004) Yes Yes – 7 QTL for chlorophyll No 76 accessions. 10/8°C for whole life, measured at 4th leaf stage. (Bano et al., 2015) Yes No No 5 The QTL on chromosome 6, probably at the end of bin 6.03, is located near to – and may be the same as – the QTL at bin 6.04 in the IBM2 2005 Neighbors 6 map, identified by Fracheboud et al. (2004). These may correspond to the luteus11 locus which affects leaf colour (Rodríguez et al., 2008). 54 (Sobkowiak et al., 2014) Chlorophyll not measured directly Yes – DEGs adjacent to QTL for chlorophyll content Yes – see above (Lee et al., 2002) Yes No No (Jompuk et al., 2005) Yes Yes – six QTL on chromosomes 1, 2, 3, 4, 10 in early-sown plants; four QTL in late-sown plants6 No (Avila et al., 2018) Yes No Differential expression of chloroplast genes under chilling stress (Rodríguez et al., 2013) Yes – lower chlorophyll and higher chlorophyllase activity in A661 Yes – QTL on chromosome 2 for chilling-induced albinism7 Yes – a putative gene in chlorophyll biosynthesis, and a chlorophyll binding protein (Revilla et al., 2016) Yes Yes – 2 SNPs for chlorophyll in chilling stress in Dent population (chromosomes 1, 4) Yes Antioxidant enzymes, or oxidative damage Association panel of 125 inbred lines. 6.4°C for 7 d at 3rd leaf stage. (Huang et al., 2013) Not measured directly No Candidate genes in 5 categories including one for antifreeze and H202 removal (Sobkowiak et al., 2014) Not measured directly Yes – DEGs adjacent to QTL related to antioxidant levels Genes for antioxidant systems identified Tolerant (S68911) and sensitive (B73) inbred lines. 14/10°C for the duration of the experiment, measured at early growth stages. (Jończyk et al., 2021) Not measured directly – but transcriptomic data suggest greater ROS scavenging in S68911 in chilling conditions No No; examined stress-response motifs and chromatin accessibility, related to chilling tolerance in the tolerant line which switched from growth to defence Abscisic acid (ABA) (Jończyk et al., 2021) Not measured directly – but transcriptomic data suggest greater ABA synthesis in tolerant line in chilling conditions No No – but see above Leaf sugar content Tolerant (S68911) and sensitive (S160) inbred Yes – decreased phloem loading in No Yes – expression of genes involved in 6 A QTL related to leaf greenness on chromosome 3 was identified as being the same as a previously identified QTL related to photosynthesis, in a population derived from the same parent lines (Fracheboud et al., 2004). Of the four QTL in late-sown plants, three were common with the early- sown plants. 7 This QTL explains 14% of phenotypic variation in chilling-induced albinism. 55 lines. 14/12°C for 28 h at 3rd leaf stage. (Bilska-Kos et al., 2016) sensitive line was observed; this leads to increased leaf sugars (not measured) phloem loading Leaf expansion (Huang et al., 2013) Yes Yes – SNPs for shoot length identified Yes – 13 genes involved in biosynthesis, metabolism, cell division and growth 56 Figures and Figure Legends Figure 1. Effects of chilling stress on maize plants. The impacts of chilling temperatures on maize physiology and morphology can be observed across a range of key traits. Growth slows down or ceases entirely, which can be observed in decreased root growth, leaf expansion and overall plant stature. The negative impact of chilling on the root system leads to decreased hydraulic conductance and partially mirrors drought stress responses, such as for example elevated abscisic acid (ABA) levels. Chilling also strongly impacts photosynthetic performance, which can be observed in decreases in CO2 assimilation, photosystem II operating efficiency (ΦPSII) and downregulation of photosynthetic genes; this can be further compounded by the accumulation of sugars due to decreased phloem loading. In addition, photoprotection via non-photochemical quenching (NPQ) is upregulated to mitigate the imbalance between light-dependent and independent reactions, but nevertheless, chilling enhances the accumulation of reactive oxygen species (ROS) as well as the breakdown of chlorophyll. Finally, chilling around the generative stages can strongly impact yield via male sterility and expansion of the anthesis-silking interval, leading to crop failure. Created with BioRender.com 57 Figure 2. Timeline of maize responses to chilling stress for nine physiological variables. Variables are grouped in three categories: photosynthetic responses in blue, photoprotective responses in orange, and signalling and developmental responses in green. Grey hatching indicates the projected time range during which the response is expected to occur, with confirmed time points indicated by coloured boxes. The darker the colour, the greater the number of studies reviewed here that reported the trend at any given time point. Studies included here do not necessarily include genetic variation, but must demonstrate the relevant response to chilling stress in at least one maize line. Many studies reveal effects following a treatment lasting the duration of the experiment, denoted by “W” for the whole experimental lifespan. NPQ: non-photochemical quenching; ABA: abscisic acid.