Ultraviolet Radiation as a Limiting Factor in Leaf Expansion and Development


Corresponding author email: j.wargent@lancaster.ac.uk (J. Wargent)


Reductions in leaf growth are a commonly observed response to ultraviolet radiation, but the underlying mechanisms remain poorly defined. This study examined the response of leaves exposed to a UV environment across a range of organizational scales, including leaf expansion rate, epidermal cell size and number, biomechanical properties, leaf–water relations and activity of cell-wall peroxidases. Two experimental approaches were used; Lettuce (Lactuca sativa L.) plants were propagated under (a) supplementary UV-B (9 kJ m−2 day−1) in controlled environment (CE) conditions, and (b) field conditions, where plants were placed under three horticultural films with differing UV transmissions. In both experiments, UV-B caused the greatest reductions in leaf expansion and final leaf size, with some reductions attributable to UV-A wavelengths. In supplementary UV-B conditions, adaxial cell size was reduced, while in field plants, both cell size and cell number were lower in an increased UV environment, as was the case with abaxial cells in CE plants. Although leaf turgor and leaf extensibility were not affected by UV wavelengths, breaking strain of leaf tissue was decreased under supplementary UV-B. Cell-wall peroxidase activity was increased in both supplementary UV conditions and in the field, where only a zero UV environment showed no upregulation of cell-wall peroxidase.


A reduction in leaf expansion is one of the most consistent responses of plant exposure to solar ultraviolet (UV) radiation (280–400 nm), and is best defined for the UV-B (280–315 nm) region of the spectrum. Supplementary UV-B irradiations both in the field and in controlled environments (CE) have been shown to consistently limit leaf expansion (1), while the attenuation of ambient solar UV-B has been shown to increase leaf expansion in a wide range of species (2,3). In addition, the more limited literature on UV-A radiation (315–400 nm) confirms that selective filtration of this waveband from sunlight also leads to increased leaf growth (4). As well as being a key driver in terms of understanding ecological and physiological responses to UV radiation, UV-mediated growth responses may also have valuable practical applications in sustainable crop production due to current legislative and socio-health pressures to reduce the use of chemical interventions in crop growth. There have been several attempts to exploit UV responses for growth regulation, using artificial UV sources (5,6), but this approach is yet to be applied in practice, partly because of concerns over gross morphological damage to crops and the practical implications of safe use (7). Another approach is the use of wavelength selective filters in horticulture, and while some progress has now been made in achieving commercially valuable regulation of crop growth under such regimes (8,9), the application of such technologies remain somewhat constrained by the still limited understanding of the underlying mechanisms of UV growth responses, particularly in field-scale, ambient UV conditions.

Leaf growth is the result of an initial phase of rapid cell division that is then followed by a period of irreversible cell expansion (10). Despite the consistent evidence that UV radiation reduces leaf expansion, the underlying mechanisms remain poorly understood, with substantial variation in the contribution of cell size and cell number reported between and within species, and between studies (11). Reductions in epidermal cell number attributable to reduced cell division have been observed in response to UV-B exposure in several species (12–15). Conversely, stimulation of cell division by UV-B has been reported in Petunia hybrida (16), with increases in leaf thickness attributed to increases in the number of spongy parenchyma cells in Brassica carinata and Medicago sativa (17). Changes in cell division in response to UV-B are considered indirect effects, since dividing leaf primordia are usually heavily shielded from UV-B (13), but the signaling system involved in such responses remains unknown. UV-B has also been shown to reduce leaf cell size in several species [e.g. (13,18–20)], although increases in cell size have been observed at lower doses (21). Cell expansion is a product of turgor and cell-wall extensibility, or the ability of the cell wall to expand or irreversibly extend during the growth cycle (22), with the rate of cell enlargement (GR) described by the equation GR = Ext(T-Y), where Ext is cell-wall extensibility, T is cell turgor, and Y is the cell wall yield threshold (23). We are unaware of any direct evidence that UV-B alters either turgor or cell-wall biomechanical properties, although changes in the latter have been inferred from biochemical responses to UV-B (15). Cell-wall extensibility is determined by cross-linkages between cellulose microfibrils which can be enzymatically strengthened (by cell-wall peroxidases) or loosened (expansins, xyloglucan endo-transglycosylases: XET) (24), and there are reports of decreases in XET activity in responses to UV-B exposure (25). Cell-wall phenolics that can act as the substrate for cell-wall peroxidase during wall stiffening have been shown to increase under UV-B (19,20,26), but there are no specific data on UV-B effects on cell-wall peroxidases, although previous studies have shown UV-increased activity of soluble peroxidases (27–29).

This study aims to examine the fundamental mechanisms through which leaf growth is affected by UV radiation. Experiments encompassed a range of organizational levels including whole plant growth analysis, biomechanical properties of the cell wall, cellular development and activity of a candidate cell wall stiffening enzyme. This was achieved using both artificial UV sources under CE conditions, and manipulation of solar (field) UV via plastic polytunnel-covering films with contrasting spectral properties, which have been developed for growth regulation in commercial crop production. Thus, experimental plants were grown under several films with contrasting UV transmissions, and analyzed accordingly. One key advantage of these complementary experimental approaches was not only to examine environmentally relevant UV-B responses, but to provide adequate UV-A, which cannot be achieved using artificial growth conditions. These two complementary experimental systems provide an integrated perspective on the processes of UV effects on leaf growth and development.

Materials and Methods

Plants and growing conditions and UV treatments—controlled environment.  Both initial propagation of plants and subsequent UV treatments were carried out in a customized CE growth room, which was composed of a custom-built UV irradiation system, allowing for the simultaneous exposure of 2 m2 of bench space to supplementary UV doses. Lactuca sativa (L.) (Iceberg-type) cv. Challenge (Syngenta Seeds Ltd., Cambridge, UK) were sown in 15 cell inserts with Levington M3 compost (Henry Alty Ltd., Preston, UK) and were propagated under a light/dark regime of 12 h/12 h at 20 ± 2°C and a Photosynthetically Active Radiation (PAR) background of 500 μmol m−2 s−1 provided from 400 W metal halide lamps (Osram HQI-BT 400 W; Osram Ltd., St Helens, UK). When the first true leaf had emerged (9 days after initial sowing), plants were exposed to one of two different UV radiation treatments as described previously (9). Briefly, background UV-A wavelengths were provided by the metal halide lamps plus 12 Q Panel UVA-340 tubes (Q-Panel Laboratory Products, Bolton, UK) filtered with clear polyester, PE (Lee Filters, Andover, UK). UV-B was provided by six UV-B tubes (Philips TL40/12-RS; Starna Ltd., Romford, UK) filtered with 0.13 mm-thick cellulose diacetate (Clarifoil, Courtaulds Ltd., Derby, UK). The zero UV-B (UVA+/UVB−) control treatment consisted of an identical setup as described above, with clear PE fitted between the UV lamp sources and the experimental plants, thus filtering out all wavelengths below 320 nm. The supplementary UV-B dose (UVA+/UVB+) was weighted using Caldwell’s generalized plant action spectrum, normalized to 300 nm (30) and the treatments provided equated to doses of zero and 9 kJ m−2 day−1. All treatments were quantified using a double scanning spectroradiometer (model SR991-v7; Macam Photometrics, Livingston, UK).

Field treatments.  Plants of Iceberg lettuce (Lactuca sativa L. cv. Challenge, Syngenta Seeds Ltd.) were raised for 14 days from sowing using a widely employed UK commercial practice at Crystal Heart Salads (Holme-on-spalding Moor, York, UK). Briefly, seeds were germinated in 4 cm3 peat blocks (Fison B2 Blocking Compost, Fisons, UK) at 16 ± 3°C in the dark for 4 days before being transferred to commercial glass for a further 10 days. At 14 days, plants were transferred to our experimental field site (Stockbridge Technology Centre: 53 N 1 W) and randomly distributed under three of the polytunnel-covering spectral filter treatments that we have previously described (9). Briefly, the film used as the control (UVA+/UVB−) is a standard commercial horticultural cladding film with transmission in the UV-B < 5%, and effectively zero below 300 nm, with transmission of at least 40% from 320–400 nm. Two films with modified UV transmission were used. The UV-opaque (UVA−/UVB−) film has transmission in the UV region of zero below 375 nm, but increasing to around 60% at 400 nm. The UV-transparent (UVA+/UVB+) film has a transmission of around 85% across the UV-B region (290–320 nm), and a transmission of at least 94% from 320 to 400 nm. All plastic cladding materials were supplied by bpi.agri Ltd. (Stockton on Tees, UK), and plants remained under the spectral filters for the propagation stage (14 days from transfer). For comparison purposes, a daily modelled irradiance for Caldwell’s generalized plant action spectrum during clear sky, summer conditions at this latitude in the UK would be approximately 4.5 kJ m−2 day−1.

Leaf area and expansion.  Daily expansion of leaf 5 (CE) and leaf 2 (field) were measured from initial leaf emergence (14 and 16 days after sowing in CE and field respectively). The day of first emergence is designated as Day 1 in all experiments. Leaf 5 was measured in CE studies and leaf 2 in field experiments because (a) these leaves had very similar growth patterns and final size under the two growing environments and (b) leaf 5 was of an adequate size for subsequent assessments of leaf biomechanical properties. Both leaf length and width measurements were taken at the widest point of the lamina using electronic digital callipers (Screwfix Direct, Yeovil, UK). Daily area growth increments (leaf length x width), were found to correlate highly with absolute leaf areas (r2 = 0.99), measured using a LI-3100A leaf area meter (LI-COR Inc., Lincoln, NE) during several destructive harvests both during and at the end of the leaf expansion phase.

Epidermal cell size/number.  At each harvest point, both during expansion (Day 8) and at the end of expansion phase (Day 18) in CE plants, and at the end of expansion phase only for field crops (Day 18), leaf 5 (CE) and leaf 2 (field) were removed in order to measure the epidermal cell size using the dental rubber impression technique (31,32). Epidermal cells were selected as a key indicator of cell size driven constraints on leaf growth as the role of the epidermis in controlling leaf growth and shape is well defined (33). The procedure involved covering the leaf surface first with dental impression material (Xantopren VL, Dental Linkline, UK). Following initial studies which indicated that there was little variation in cell size and cell number across the majority of the leaf surface, dental impression gel was applied at the central region of the lamina (adaxial surface only for field plants), avoiding the mid-vein. Once the material had set (1–2 min), the leaf was peeled away. Acrylic-based varnish was used to produce a translucent positive replica from the negative rubber impression. Images of the epidermal impressions were taken at 100× magnification with a light microscope fitted with a CCD digital camera (Spot Insight; Diagnostic Instruments Inc., MI), with cell size and cell number per unit area analyzed using Image Pro Plus v4.5 (Media Cybernetics Inc., MD). In order to calculate the number of epidermal cells per leaf, final leaf area of the imprint leaves was also determined at the point of harvest as described above.

Leaf biomechanical properties.  On Day 7 (during-expansion phase) and 17 (post-expansion phase), biomechanical properties of CE plants were measured by carrying out tensile tests on leaf segments using a Universal Testing Machine (model 4301; Instron, High Wycombe, UK). Rectangular leaf strips (10 × 5 mm) were cut from the central region of the lamina, parallel to and avoiding the mid-vein and their thickness measured with calipers. Leaf samples were then immediately clamped into the jaws of the Instron, being attached using cyanoacrylate glue between strips of paper, and stretched at a rate of 10 mm s−1. An interfaced computer simultaneously recorded a readout of force required vs. extension and calculated the tensile properties of the leaf material. The breaking stress or strength was equal to the breaking force divided by the cross-sectional area of the leaf sample. The breaking strain was equal to the extension at which the leaf broke divided by its original length, and the extensibility or Young’s modulus was the initial slope of the stress/strain curve (34,35).

Leaf turgor—psychrometry.  Leaf turgor in CE plants was assessed on Day 8 and 18 (during and post-expansion) using dew-point psychrometry to assess both leaf water (ΨW) and osmotic (ΨΠ) potential (36,37). While assessments of leaf water potential by thermocouple psychrometry are typically affected by continued growth during the incubation phase (38,39), leaf water potential quantification using the rapid pressure chamber method was deemed an unsuitable approach in this study given the morphology of L. sativa. Similarly, the use of the micro-pressure probe technique proved unsuitable due to the likelihood of tip clogging effects caused by the latex exudate of L. sativa. Circular discs (0.5 cm2) from leaf 5 were cut from the same region of the lamina as used for previous analyses, and were then placed into C52 sample chambers (Westcor, Logan, UT) connected to a HR-33T microvoltmeter (Westcor) maintained at 25°C. Voltage was then recorded after a predetermined 3 h period equilibration period in order to reach thermodynamic equilibrium, and converted to MPa (ΨW) based on a calibration curve of NaCl solutions of known ΨW and molarity. Osmotic potential was then quantified by wrapping the leaf discs individually in aluminum foil, which were then plunged into liquid N2 for 10 s in order for cell lysis to occur. After allowing the discs to thaw to 25°C, leaf discs were placed back into the sample chambers, this time with an equilibration period of 30 min, with voltage recorded once again and converted to MPa, allowing for the calculation of leaf turgor (ΨP = ΨW–ΨΠ).

Extraction and assay of cell-wall peroxidase activity.  On the first day following emergence, and for four subsequent days for CE plants (5 days for field plants), leaf 2 (in both field and CE plants) was removed and fresh weights were determined for each sample. These were then stored in liquid N2 for subsequent assay of cell-wall peroxidase activity. Extraction of cell-wall peroxidase was carried out using the method outlined by Bacon et al. (40). Briefly, samples were homogenized in liquid N2 before being transferred to ice-cold buffer (10 mm sodium succinate, 10 mm calcium chloride, 1 mm dithiothreitol; Sigma–Aldrich, Poole, UK), pH 6.0, at a ratio of 10:1 buffer to sample fresh weight. 200 μL of buffer and homogenized tissue was then centrifuged at 2000 g for 5 min. The pellet was washed four times in 200 μL of 10 mm sodium succinate (pH 6.0) to remove soluble peroxidase activity, before being re-suspended in an equal volume of final extraction buffer (50 mm sodium succinate, 1 m sodium chloride, pH 6.0) to extract activity from the cell wall. The four low ionic concentration buffer wash stages reduced the activity derived from tissue by ca. 95% (data not shown). The final high salt wash extract showed an increase in the activity when compared with the final wash of the four low salt concentration washes. Activity was determined by assaying a 100 μL sample of the supernatant using the Guaiacol method detailed by Chance and Maehly (41). Briefly, the 100 μL sample was added to 1 mL of 20 mm sodium phosphate buffer, which contained 276 μL of Guaiacol (Sigma–Aldrich, Dorset, UK) per 50 mL of buffer. The reaction was started by adding 100 μL of 0.03% hydrogen peroxide in distilled water (w/w) (Sigma–Aldrich). The concentration of hydrogen peroxide and Guaiacol used gave a linear change in absorbency over 20 + min. This reaction was mixed in 1.5 mL spectrophotometric cuvettes with the absorbency of the solutions at 470 nm then measured after 20 min at 25°C.

Statistics.  Field studies were repeated over three UK summer growing seasons during the period 2004–2007 where consistent results were observed and field data presented here are from a single representative field experiment. For both CE and field experiments there were 10–15 plants per treatment, with single time-point comparisons between treatments analyzed by t-test, timecourse analyses between treatments analyzed by one-way and two-way repeated measures analysis of variance, and regression analyses for cell-wall peroxidase/leaf expansion relationships by analysis of variance F-test using spss v 11.5 (SPSS Inc., Chicago, IL).


Leaf area and expansion

In CE conditions, leaf 5 area was significantly reduced in UVA+/UVB+ plants after 5 days of exposure to the supplementary UV-B treatment, with UVA+/UVB+ leaves 50% smaller (< 0.001; Fig. 1A). By the end of the expansion phase (Day 17), there was a 20% UVB-mediated reduction in leaf size (Fig. 1A), equating to a total reduction of 988 mm2 in overall leaf 5 size (< 0.05; Fig. 1A). In the field treatments, the UVA+/UVB+ treatment showed a significant decrease in the size of leaf 2 by the end of the expansion period compared to both UVA+/UVB− and UVA−/UVB− environments (< 0.001; Fig. 1B). In terms of the three films used in the field, the only significant differences are between the UVA+/UVB+ and UVA−/UVB− treatments (< 0.05; Fig. 1B), with the UVA+/UVB- treatment showing consistent but nonsignificant reductions in leaf size when compared with the UVA−/UVB− film.

Figure 1.

 Effects of UV on cumulative leaf expansion in: (A) controlled environment growth conditions: supplementary UV-B (9 kJ m−2 day−1 plant weighted UV, UVA+/UVB+ = •) and zero UV-B (UVA+/UVB− = ○), and (B) field: UV-B and UV-A (UVA+/UVB+ = •), zero UV-B and reduced UV-A (UVA+/UVB− = ○), and zero UV (UVA−/UVB− = bsl00072). ±SE, = 10 (CE), 15 (field). Different letters indicate a significant difference between UV treatments across all days (< 0.05).

When these changes in leaf size are expressed in terms of daily leaf expansion rate (LER mm2 day−1), all plants reached the point of maximum LER approximately halfway through the expansion phase (Day 8 – CE, Day 7 – field; Fig. 2). In plants exposed to either supplementary UV-B in CE conditions or the UVA+/UVB+ field treatment LER was reduced from early in expansion until just after the point of maximum LER, with the largest differences between treatments observed at the point of maximum daily expansion rate. At that point, there was a 43% reduction under increased UV-B in the CE) studies (< 0.001; Fig. 2A), and in the field there was a 31% and 33% significant reduction of the UVA+/UVB+ leaves when compared with both UVA+/UVB− and UVA−/UVB− respectively (both < 0.01; Fig. 2B).

Figure 2.

 Effects of UV on daily leaf expansion rate in: (A) controlled environment growth conditions: supplementary UV-B (9 kJ m−2 day−1 plant weighted UV, UVA+/UVB+ = •) and zero UV-B (UVA+/UVB− = ○), and (B) field: UV-B and UV-A (UVA+/UVB+ = •), zero UV-B and reduced UV-A (UVA+/UVB− = ○), and zero UV (UVA−/UVB− = bsl00072). ±SE, = 10 (CE), 15 (field). Symbols denote significant differences at < 0.001 (***) and < 0.01 (**) for UVA+/UVB+ treatments in CE and field conditions respectively at point of maximum expansion.

Epidermal cell size and number

In CE experiments, adaxial epidermal cell area was significantly reduced in UVA+/UVB+ plants compared with UVA+/UVB− (< 0.001; Fig. 3A), but numbers of adaxial epidermal cells per leaf did not change significantly with treatment (Fig. 3A), 89% of the leaf area reduction was accounted for by reduced cell size and 11% of the variation by cell number (Fig. 3A). Under the field treatments, much the same was observed in terms of adaxial cell size, with UVA+/UVB+ plants showing significantly reduced epidermal cell areas compared with UVA−/UVB− (< 0.001; Fig. 3B) but not when compared with UVA+/UVB− (Fig. 3B). However, in contrast to the UVB-supplemented CE plants, epidermal cell number per leaf from the field treatments was significantly lower in UVA+/UVB+ leaves compared with UVA−/UVB− (< 0.05; Fig. 3B) and UVA+/UVB− (< 0.05; 3B), with no differences between UVA+/UVB+ and UVA+/UVB− treatments. In CE experiments, the abaxial leaf surface was also measured. The relative balance between changes in cell size and cell number were different than that of the upper surface, with both cell size and cell number significantly decreased in UVA+/UVB+ leaves (< 0.05; Fig. 3C). In contrast to the adaxial surface of UVA+/UVB+ leaves, 46% of the UV-mediated reduction in leaf area was explained by a reduction in cell size, as opposed to 54% of the difference attributed to fewer cells.

Figure 3.

 Effects of UV on epidermal cell development. Number of epidermal cells per leaf (closed bars) and epidermal cell area (gray bars); (A) adaxial surface in CE plants: supplementary UV-B (9 kJ m−2 day−1 plant weighted UV, UVA+/UVB+) and zero UV-B (UVA+/UVB−), (B) adaxial surface in field plants: UV-B and UV-A (UVA+/UVB+), zero UV-B and reduced UV-A (UVA+/UVB−), and zero UV (UVA−/UVB−), (C) abaxial surface in CE plants: UV-B (UVA+/UVB+) and zero UV-B (UVA+/UVB−). ±SE, = 10 (CE), 15 (field). Different letters indicate a significant difference between UV treatments (< 0.05).

Leaf water relations and leaf biophysical properties

In the biophysical assessments, the force required to stretch leaf samples rose linearly with extension at first, before reaching a yield point after which the leaf broke. Breaking strain in UVA+/UVB+ plants during the expansion phase (Day 7) was significantly lower than that in UVA+/UVB− plants (< 0.05; Table 1) as was breaking stress required (< 0.01; Table 1), but there were no significant differences in leaf extensibility or leaf thickness. By the end of the expansion phase at Day 17, there were no significant differences between treatments. There were no significant differences in leaf water potential, osmotic potential or leaf turgor between treatments in either during or post-expansion phase timepoints (data not shown).

Table 1.   Biophysical characteristics of leaf 5 as leaf thickness (mm), breaking stress (MPa), % extensibility, and breaking strain (mm/mm) according to UV treatment during leaf expansion phase (Day 7) and post-expansion phase (Day 17); supplementary UV-B (9 kJ m−2 day−1 plant weighted UV, UVA/UVB+) and zero UV-B (UVA+/UVB−) = 15.
 Leaf thickness (mm)Breaking stress (MPa)Extensibility (MPa)Breaking strain (mm mm−1)
  1. P-values are shown, with asterisks indicating a significant difference between UV treatments (*< 0.05, **< 0.01).

During expansion

Cell-wall peroxidase activity

UVA+/UVB+ leaves showed a significant increase in cell-wall peroxidase activity over the 5 day harvest period (< 0.05; Fig. 4A), with the most marked increase compared with UVA+/UVB− leaves on Day 3, prior to the onset of UVB-mediated reductions in leaf size. However, the UVA+/UVB− treatment did show cell-wall peroxidase activity significantly greater than zero (< 0.001; Fig. 4A) with a similar increase in enzyme activity over the course of the study as seen in UVA+/UVB+. In the field, the UVA+/UVB+ treatment exhibited significantly increased cell-wall peroxidase activity compared with both the UVA+/UVB− (< 0.05; Fig. 4B) and UVA−/UVB− treatments (< 0.001; Fig. 4B). UVA+/UVB− leaves also showed significant increases in activity throughout the harvest period when compared with the UVA−/UVB− treatment (< 0.001; Fig. 4B). Cell-wall peroxidase activity remained extremely low in UVA−/UVB− plants throughout the experiment (Fig. 4B). In a similar manner to the CE trials, both the UVA+/UVB− and UVA+/UVB+ filter treatments exhibited the greatest increase in activity relative to the UVA−/UVB− treatment on Day 3 (< 0.05; Fig. 4B) and Day 4 respectively (< 0.01; Fig. 4B), immediately prior to the onset of clear differences in leaf growth. The relationship between changes in cell-wall peroxidase and changes in leaf growth was then subsequently examined by comparing percentage decrease in daily LER against the percentage increases in cell-wall peroxidase activity expressed relative to both UVA+/UVB− plants (field), or UVA+/UVB− plants (CE). There were consistently significant negative correlations between changes in cell-wall peroxidase and changes in daily LER (data not shown), with a high correlation coefficient observed when a delay of 2 days between cell-wall peroxidase activity and daily LER was imposed (r2 = 0.794, < 0.01 for the relationship across all treatments).

Figure 4.

 Cell-wall peroxidase content (expressed per nmol tetraguaiacol mg−1 FW h−1) according to UV treatment; (A) CE conditions: supplementary UV-B (9 kJ m−2 day−1 plant weighted UV, UVA+/UVB+ = •) and zero UV-B (UVA+/UVB− = ○), (B) Field: UV-B and UV-A (UVA+/UVB+ = •), zero UV-B and reduced UV-A (UVA+/UVB− = ○), and zero UV (UVA−/UVB− = bsl00072). ±SE, = 10 (CE), 15 (field). Different letters indicate a significant difference between UV treatments across all days (< 0.05).


While there is a strong consensus that reductions in leaf growth are a significant response to UV radiation across a wide range of species (1,42), no study to our knowledge has elucidated several key aspects of the underlying ecophysiology in one experimental system. This study has examined the mechanistic basis for the inhibition of leaf growth by ultraviolet radiation across several scales, and the nature of such physiological responses has been observed at the field-scale, as well as under CE conditions.

Despite the fact that both field and CE studies consistently showed significantly reduced cell expansion in response to UV radiation, responses in terms of cell number were more variable. While there is typically some variability in epidermal cell size across the leaf surface of many species, particularly towards the edges of the lamina, and although changes in adaxial cell number were not significant in the CE experiments, in the field there were clear reductions in adaxial cell number as well as cell size in response to increasing UV. Contrasts between CE and field are not surprising since treatments were not exactly comparable, notably the spectral filters in the field manipulated UV-A as well as UV-B. Although the quantification of unweighted UV-A does little to describe the differences in UV fluxes between CE conditions and the field, the newer ‘UV plant growth weighting function’ (UVFlint and Caldwell) (43) showed increased sensitivity into the UV-A region compared with the Caldwell weighting function (UVCaldwell) (29). So while the UV dose in CE conditions was quantified as 9 kJ m−2 day−1 using UVCaldwell, the UVFlint and Caldwell weighted dose = 19 kJ m−2 day−1. Such comparisons make for an interesting contrast with field conditions, where modeled UVCaldwell under clear summer conditions is approximately 4.5 kJ m−2 day−1 (0.5× CE dose), whereas UVFlint and Caldwell would be expected to exceed 24 kJ m−2 day−1 (1.26× CE dose) (9). It is possible that UV-A has a greater effect on cell proliferation than UV-B, possibly because of responses initiated by the differing photoreceptors for the two wavebands, e.g. cryptochrome-mediated growth responses to UV-A (44). The contrasting balance between the effects of UV-B on cell number and cell size between adaxial and abaxial epidermises in CE conditions is particularly interesting. However, using UV-B doses far exceeding UK ambient fluxes (32 kJ), Nogues et al. (15) reported similar contrasts between different leaf layers in pea (Pisum sativum). Such differences might reflect the penetration of UV into leaves, particularly the very limited penetration of UV-B to the abaxial leaf surface (45). Although differences between upper and lower epidermises could reflect responses to the immediate UV environment of specific cell-layers, this is unlikely for cell division, since proliferation is largely complete before leaf expansion, and hence before direct UV-exposure (13). Alternatively, the observed differences could reflect coordinated plant responses to UV exposure. There is increasing recognition that UV-B responses should be seen as the combination of UV-specific effects and more generalized plant responses to their environment (46), and this may be the case in our study. General responses might be analogous to the compensation that occurs in response to changes in cell volume and/or cell number during differing phases of leaf growth (47). There is certainly evidence of such compensatory mechanisms and their influence on leaf development, e.g. the role of ASYMMETRIC LEAVES1 and 2 proteins in the control of adaxial and abaxial leaf patterning (for review, see Hay et al. (48), in addition to the possible role of hormones such as auxin in promoting cell fate (49) and other factors influencing cellular development such as endopolyploidy (50).

This model of reduced leaf expansion as a coordinated response of cell division and cell expansion to UV is a very different model than direct photochemical damage. The variation in the balance between changes in cell division and cell expansion seen in previous studies of UV is more consistent with such co-ordination (with different plants acclimating to different conditions in different ways) than with a single “driver” of reduced leaf expansion in response to UV-B. The well-known tendency of leaves to “cup” under high UV-B doses (51) might also reflect the limits on such co-ordination. The reductions in LER in both UVA+/UVB+ plants under CE conditions and both UVA+/UVB+ and UVA+/UVB− plants from the field treatments were consistently associated with reductions in epidermal cell expansion. Despite previous suggestions that changes in turgor pressure could be a key mechanism mediating UV-B-induced reductions in cell expansion (15), we found no evidence for this here. In addition, we observed no effects of UV-B on cell-wall extensibility, despite the fact that decreases in cell-wall extensibility have frequently been associated with decreases in LER during typical growth (52,53). Therefore, it is possible that UV-B-mediated growth inhibition could be as a result of changes in cell wall yield threshold (Y), defined as the turgor which exceeds a minimum threshold in order for cell wall extension to occur (23). While there is currently no direct evidence to support this, there is evidence that cell wall yield threshold can be affected by age of leaf and relative humidity (38), in addition to drought conditions (54). In this experiment, significant decreases were observed in the breaking strain (and stress) of leaf tissue. While such decreases in breaking strain of leaf samples are somewhat difficult to interpret in the absence of observed changes in mechanical stiffness, it is likely that when compared to the high strain rates applied in this study, the limits to plastic extension in growing leaves may be reduced under UV-B (55), suggesting higher breaking strains under typical growth, and more plastic deformation to the visoelastic cell walls, which may have implications for crop resistance to mechanical damage that is inherent in intensive cropping regimes. In addition, such findings may suggest that UV-B affects the number of cross-links between microfibrils or increased stiffness of the matrix of the cell wall rather than the number of cellulose microfibrils in the cell wall or their properties (56,57). This is consistent with the upregulation of cell-wall peroxidase under increased by UV radiation (demonstrated here both in CE and field experiments), since the role of cell-wall peroxidases in upregulating cross-linkages between components of the cell wall is well established (23). It was notable that the field data also showed clear effects of longer wavelength UV-A wavelengths of cell-wall peroxidases, since only the zero-UV field environment showed no upregulation of peroxidase activity. Reductions in LER under increased UV-B (21,58) and UV-A (4) have been reported, but our data suggests both are related to induction of cell-wall peroxidase. The role of cell-wall peroxidases in regulating cell expansion is well documented (59,60), and similar observations of growth inhibition following induction of cell-wall peroxidase have been noted in response to herbivory (61) and in the induction of systemic resistance (62), and it is likely that induction of peroxidase may regulate growth following pathogen attack (63).

While the UV-mediated mechanisms controlling processes such as leaf expansion may now be becoming better understood, our knowledge of how such responses combine with other related systems such as secondary metabolism and UV-signaling, forming part of a wider multiple plant-stress response network are still in their infancy, and this is a key area of focus for the future, particularly in the context of exploiting such responses to increase sustainability in crop growth, in addition to the related areas of integrated pest and disease control. Greater understanding of the relationships between mechanisms of UV-B and UV-A-mediated growth responses is another important aspect of understanding the total UV whole-plant response, and as such is worthy of further investigation in a similar manner to the attention now being given to UV-dependant molecular systems of response.


Acknowledgements— We are grateful to the Horticultural Development Council for funding this work with a studentship to J.J.W. (CP 26) and a Council Fellowship to J.P.M. (PC 221). Our thanks also go to various technical staff at Lancaster University Department of Biological Sciences, Stockbridge Technology Centre, bpi-agri for the supply of spectral filters, James Bean (Crystal Heart Salads) for grateful donations of plant material, and John Doonan (John Innes Centre) for helpful comments on the manuscript.