Photosynthesis Research

, Volume 125, Issue 1, pp 291–303

Cadmium accumulation in chloroplasts and its impact on chloroplastic processes in barley and maize

  • Eugene A. Lysenko
  • Alexander A. Klaus
  • Natallia L. Pshybytko
  • Victor V. Kusnetsov
Regular Paper

DOI: 10.1007/s11120-014-0047-z

Cite this article as:
Lysenko, E.A., Klaus, A.A., Pshybytko, N.L. et al. Photosynth Res (2015) 125: 291. doi:10.1007/s11120-014-0047-z


Data on cadmium accumulation in chloroplasts of terrestrial plants are scarce and contradictory. We introduced CdSO4 in hydroponic media to the final concentrations 80 and 250 μM and studied the accumulation of Cd in chloroplasts of Hordeum vulgare and Zea mays. Barley accumulated more Cd in the chloroplasts as compared to maize, whereas in the leaves cadmium accumulation was higher in maize. The cadmium content in the chloroplasts of two species varied from 49 to 171 ng Cd/mg chlorophyll, which corresponds to one Cd atom per 728–2,540 chlorophyll molecules. Therefore, Mg2+ can be substituted by Cd2+ in a negligible amount of antenna chlorophylls only. The percentage of chloroplastic cadmium can be estimated as 0.21–1.32 % of all the Cd in a leaf. Photochemistry (Fv/Fm, ΦPSII, qP) was not influenced by Cd. Non-photochemical quenching of chlorophyll-excited state (NPQ) was greatly reduced in barley but not in maize. The decrease in NPQ was due to its fast relaxing component; the slow relaxing component rose slightly. In chloroplasts, Cd did not affect mRNA levels, but content of some photosynthetic proteins was reduced: slightly in the leaves of barley and heavily in the leaves of maize. In all analyzed C3-species, the effect of Cd on the content of photosynthetic proteins was mild or absent. This is most likely the first evidence of severe reduction of photosynthetic proteins in leaves of a Cd-treated C4-plant.


Cadmium Chloroplasts mRNA Proteins Barley Maize Photosystem II activity 


BEP clade

Main branch of Poaceae family that includes subfamilies Bambusoideae, Ehrhatoideae, Pooideae






Coefficient of electron-transport rate


Dry weight


Minimum Chl a fluorescence in the dark-adapted state


Maximum Chl a fluorescence


Maximum quantum yield of PSII


Fresh weight


Heavy metal


Large or small subunits of Rubisco


Coefficient of non-photochemical quenching of excited Chl state

PACMAD clade

Main branch of Poaceae family that includes subfamilies Panicoideae, Aristidoideae, Chloridoideae, Micrairoideae, Arundinideae, Danthonioideae


Phosphate buffer saline


Phosphoenolpyruvate carboxylase


Photosystems I


Photosystems II


Fast relaxing component of NPQ (energy/ΔpH dependent)


Slow relaxing component of NPQ (dependent mainly on photoinhibition)


Coefficient of photochemical quenching of excited Chl state


Coefficient of non-photochemical quenching of excited Chl state


Sodium dodecylsulfate polyacrylamide gel electrophoresis


Effective quantum yield of PSII


Cadmium disturbs many processes in plants. Photosynthesis is one of the most important targets that should be protected from cadmium action. In vitro high concentrations of the heavy metal (HM) very rapidly decrease functional activity of photosystem II (PSII) but not photosystem I (PSI) in isolated chloroplasts (e.g., Bazzaz and Govindjee 1974). “Each plant organ, due to membrane selectivity and mechanisms of heavy metals immobilization, is a barrier to Cd movement from the soil to the chloroplasts” (Siedlecka and Krupa 1999). The vast majority of plant species restrict Cd penetration from a soil to aerial organs (Baker 1981). In a leaf, Cd is absorbed by cell wall polysaccharides or transported into vacuoles; some plants excrete cadmium into trichomes or on the leaf surface (Hagemeyer and Waisel 1988; Salt et al. 1995; Choi et al. 2001). However, there remains a small portion of the cadmium that invades the chloroplasts, which raises the question—how big is the portion?

In a dozen works, roughly prepared chloroplast fractions were analyzed. In such a fraction, the cadmium content may be overestimated due to contamination by organelles with high cadmium content (e.g., cell walls, vacuoles), or it may be underestimated due to loss of intactness by analyzed chloroplasts. The best way to obtain highly purified intact chloroplasts is centrifugation in a density gradient. Such an approach was employed in a few articles discussed below. Chloroplasts of the algae Chlamidomonas reinhardtii (Nagel et al. 1996) and Euglena gracilis (Mendoza-Cozalt et al. 2002) were able to accumulate more than 50 % of the cellular cadmium. However, algae vary greatly in cell organization and therefore in cadmium compartmentalization. For example, Euglena cells lack cell walls and vacuoles and use their enormous chloroplast as some sort of vacuole. Thus, approximately 60 % of the Cd accumulated in their cells was found in chloroplasts (Mendoza-Cozalt et al. 2002). In contrast, the red algae Audouinella saviana (Meneghini) Woelkerling has both cell walls and vacuoles, so no Cd signal was detected from their chloroplasts by X-ray electron microscopy (Talarico 2002).

Two groups analyzed Cd accumulation in chloroplasts of the dicotyledonous plant Brassica napus (Baryla et al. 2001) and the monocotyledonous species Phragmites australis (Pietrini et al. 2003). Both groups tested cadmium action for a long enough period (21 and 47 days), but they used very different species and different experimental designs. Hence, the plants showed no similarity in Cd accumulation in leaves or in chloroplasts. Chloroplasts accumulated drastically different portions of the leaf cadmium: 0.02 % in rape (Baryla et al. 2001) and 10–15 % in common reed (Pietrini et al. 2003). We failed to find any other work dealing with Cd content in highly purified chloroplasts. Thus, the magnitude of Cd accumulation in plant chloroplasts remains obscure.

Cadmium can influence photosynthetic processes by indirect action. It lowers stomatal density, stomatal conductance, and CO2 uptake, which in turn limits photosynthesis (Baryla et al. 2001). The presence of cadmium reduces the chloroplast content both per cell and per unit of leaf area (Baryla et al. 2001; Fagioni et al. 2009). Therefore, the net CO2 assimilation rate may be reduced when calculated on a leaf area basis but unchanged when calculated per chlorophyll unit (Pietrini et al. 2003). Also, Cd inhibits Fe uptake by roots, and Fe deficiency leads to leaf chlorosis (Siedlecka and Krupa 1999).

Existing data about the effect of Cd on photosynthesis are contradictory. Processes in the electron-transport chain of the thylakoid membranes—potential (Fv/Fm) and effective (ΦPSII) quantum yield of PSII photochemistry, photochemical (qP) and non-photochemical (qN, NPQ) quenching of the excited state of chlorophylls—may be influenced by cadmium. This has been well documented for both monocots (He et al. 2008; Cai et al. 2011) and dicots (Pietrini et al. 2010; Filek et al. 2010), and at least some of these parameters were reduced in hyperaccumulator species (Küpper et al. 2007). However, a number of articles have demonstrated that these parameters were not changed in monocotyledonous species (Drazkiewicz et al. 2003; Wu et al. 2003; Ekmekci et al. 2008), dicotyledonous species (Baryla et al. 2001; Burzynski and Zurek 2007), and hyperaccumulator species (Zhou and Qiu 2005; Tang et al. 2013) that were grown in the presence of cadmium. The distinction can be found among the cultivars of one species. For example, long-term exposure to 5 μM Cd reduced the potential and actual quantum yield of PSII in two barley cultivars and didn’t influence these parameters in another two cultivars of the species (Wu et al. 2003). Short-term exposure of maize seedlings to very high concentrations of Cd (300 or 600 μM) decreased the parameters, mainly reflecting photochemical activity of PSII, (Fv/Fm, ΦPSII, qP, ETR) in a sensitive line but not in a tolerant ones (Ekmekci et al. 2008).

These discrepancies can be explained at least partially by different levels of cadmium intrusion into chloroplasts. However, the existing data are insufficient. Therefore, the present research was focused on the comparison of cadmium accumulation in chloroplasts with its impact on chloroplastic processes. To discriminate between species-specific and more general effects, we used two species belonging to basically diverged branches of the Poaceae family: Hordeum vulgare (BEP clade) and Zea mays (PACMAD clade). Poaceae species employ advanced mechanism of Fe uptake that is not inhibited by Cd (Siedlecka and Krupa 1999), and cereals are less vulnerable to cadmium toxicity (Inouhe et al. 1994) and Cd-induced chlorosis (Siedlecka and Krupa 1999). Additionally, barley and maize have different types of photosynthesis (C3 and C4, respectively). We analyzed the toxic effect of cadmium on seedling growth, on accumulation of Cd in organs and chloroplasts, on the functional activity of PSII (Fv/Fm, ΦPSII, qP, qN, NPQ), and on the steady-state levels of 10 chloroplast mRNAs and two chloroplast proteins.


Plant growth conditions

Maize (Z. mays L. cv. Luchistaya) and barley (H. vulgare L. cv. Luch) seedlings were grown at 180–200 μmol photons m−2 s−1 and a photoperiod of 16 h light/8 h dark under continuous aeration on modified Hoagland medium containing 3 mM KNO3, 2 mM Ca(NO3)2, 1 mM NH4H2PO4, 0.5 mM FeSO4, 0.5 mM MgSO4, 25 μM H3BO3, 2 μM ZnSO4, 2 μM MnSO4, 1 μM KCl, 0.1 μM CuSO4, 0.1 μM (NH4)6Mo7O24, 2 mM MES, pH 6.5. The temperature was optimal for each species: 25 °C for maize and 21 °C for barley. A heavy metal (cadmium, copper, or nickel) was added as a sulfate.

Caryopses were immersed in 0.25 M CaCl2 for 30 min and then kept for 2 days at 4 °C in darkness on filter paper moistened with 0.25 M CaCl2. Imbibed caryopses were germinated for 2 days under the same conditions but at 21/25 °C. Plant age was determined from the start of germination at 21/25 °C. Germinated caryopses were transferred to vessels for plant growing containing Hoagland medium. A HM was added the next day just before coleoptile protrusion. All HMs were introduced in hydroponic media from aqueous stock solution (100 mM). CdSO4 was added to the final concentrations 4, 20, 80, 100, 200, or 250 μM. CuSO4 or NiSO4 was added to the final concentration 100 μM. The analyses were performed on 9-day-old seedlings exposed to heavy metal for 6 days. All experiments were repeated at least three times with two exceptions: the biological experiment for Western blotting was repeated twice, with very similar results obtained, and cadmium accumulation in chloroplasts of maize at 250 μM Cd was measured in two biological experiments only.

For the assessment of lethal effect, cadmium concentration was increased to 250 μM, and the exposure period was extended to 31 days. Barley seedlings were cultivated as 26 per a 1.2 l vessel. In four vessels, cadmium was added as CdSO4, and in other four it was added as CdCl2. Hoagland media with Cd were renewed once a week. Cadmium plants were considered dead when they were completely dry and had no spots of green tissue. The effects of both salts were very similar, and all the data were united in the single set (N = 208).

Plant height was determined as the distance between caryopses and the highest shoot point. The area of the leaf blade was calculated according to the equation: S = 2/3 Ld, where L is the length of the leaf blade and d is the leaf width in the middle of the blade (Anikeev and Kutuzov 1961). For determination of water content, tissues were dried at 60 °C overnight.

Chlorophyll estimation

The content of photosynthetic pigments was determined in leaf samples and in isolated chloroplasts. In a leaf blade, the terminal 1-cm segment was removed, and the following 1-cm segment was analyzed. Pigments were extracted with 80 % acetone, and concentrations of chlorophylls (Chl a and Chl b) and carotenoids (Car) were estimated according to (Lichtenthaler 1987).

Cadmium measurement

For determination of cadmium content, tissues were dried at 60 °C overnight. Roots were initially washed in distilled water for 10 min. Dried samples (50–100 μg) or chloroplast pellets were incubated overnight in a solution containing 1.5 mL of 64 % nitric acid and 0.6 mL of 65 % perchloric acid. The next day, samples were heated for 1.5 h at 150 °C, then for 2 h at 180 °C; thereafter, 50 μL of concentrated hydrogen peroxide was added, and samples were left overnight. Afterward, samples were adjusted to the final volume (10 mL) with distilled water. Cadmium concentration was determined with a Formula FM400 atomic-absorption spectrophotometer (Labist, Russia).

Chlorophyll fluorescence

Chl a fluorescence was measured at room temperature with a PAM 201 fluorometer (Walz, Germany). Before measurement, samples were adapted to darkness for 15 min. Measured light (650 nm, 0.04 μmol photons m−2 s−1) modulated with low frequency (8 kG) excited the initial fluorescence level F0. A light 1-s pulse (3,500 μmol photons m−2 s−1) caused fluorescence to increase to its maximal level Fm; the maximum quantum yield of PSII was calculated as Fv/Fm, where Fv = Fm − F0. The effective quantum yield of photosystem II (PSII) photochemistry (ΦPSII) was calculated on the basis of slow Chl a fluorescence kinetics, and it was measured with actinic light (665 nm, 480 μmol photons m−2 s−1) and periodical light pulses (3,500 μmol photons m−2 s−1) using the following equation: \(\varPhi {\text{PSII}} = \left( {F^{\prime}_{\text{m}} - F} \right)/F^{\prime}_{\text{m}}\). Photochemical (qP) and non-photochemical (qN, NPQ) coefficients of quenching was calculated as \(qP = \left( {F^{\prime}_{\text{m}} - F} \right)/\left( {F^{\prime}_{\text{m}} - F^{\prime}_{0} } \right),\)\(qN = \left( {F_{\text{m}} - F^{\prime}_{\text{m}} } \right)/ \left( {F_{\text{m}} - F_{0} } \right),\)\({\text{NPQ}} = \left( {F_{\text{m}} /F^{\prime}_{\text{m}} } \right) - 1 ,\) (Krause and Weis 1991; Lichtenthaler et al. 2005).

The dark NPQ relaxation kinetics were measured after 15 min irradiation of dark-adapted seedlings by red actinic light (480 μmol photons m−2 s−1). The magnitude and half-time of different kinetic components of NPQ relaxation kinetics were obtained from plots of log (Fv) as a function of relaxation time. A best linear fit was calculated for each kinetic component providing its half-time as the time at which NPQi is 50 % of the maximum magnitude, which is obtained from the intersection of the linear fit with the ordinate (Horton and Hague 1988; Walters and Horton 1991).

Chloroplast isolation

All procedures were performed at 4 °C in shaded light. For the analysis of cadmium content in chloroplasts, all equipment (mainly glass) was presoaked thrice in 2 mM EDTA and once in 0.1 N HCl for at least 2 h each time. Leaves (10–15 g) were homogenized in 60 mL of buffer A (50 mM Tris–HCl, pH 8.0, 0.4 M sorbitol, 15 mM NaCl, 2 mM EDTA, 5 mM β-mercaptoethanol). The homogenate was filtered through one layer of cheesecloth and two layers of miracloth (Calbiochem–Behring, United States) and centrifuged in a K-23 centrifuge for 1 min at 3,000 rpm (1,400×g at the bottom of the tube). The organellar pellet was gently resuspended in 15 mL of buffer A and fractionated in a three-step (20/40/70 %) Percoll (GE Healthcare, United States) gradient in a K-23 centrifuge for 10 min at 5,500 rpm (4,000×g at the fractionating level). Intact chloroplasts were collected at the layer interface and washed by adding buffer A to a total volume of 1.5 mL, with subsequent centrifugation at 3,800 rpm for 1 min (Heraeus Biofuge fresco centrifuge, 4 °C). The supernatant was carefully removed with an automated pipette. For RNA analysis, chloroplasts were collected at the 40/70 % interface. However, this fraction was small and insufficient to estimate Cd content. The bulk of the chloroplasts was collected at the 20/40 % interface and used for Cd analysis.


Chloroplasts were dissolved in buffer B (20 mM Tris–HCl, pH 7.0, 4 M guanidine thiocyanate, 20 mM EDTA, 0.7 % lauroyl sarcosin, 100 mM β-mercaptoethanol), and RNA was isolated according to standard method described in (Lysenko et al. 2013). Residual DNA was removed by DNase I (Thermo Scientific) treatment.

The first cDNA strand was synthesized with reverse transcriptases RevertAid or RevertAid Premium (Thermo Scientific) using gene-specific anti-sense primers; for each gene, cDNA was generated separately. For the synthesis, 100 ng of chloroplast RNA per 10 μL of reaction mixture was used. Reverse transcription was performed at 42 °C (RevertAid, genes petL, atpA, atpB, atpF, ndhA, cemA), 50 °C (RevertAid Premium, genes petB, petD, ndhB), or 55 °C (RevertAid Premium, gene ndhF).

For PCR, 2 μL of cDNA and 1.25 units of DNA polymerase were added per 20 μL of reaction mixture. Taq DNA polymerase (SibEnzyme, Russia) was used for genes petL, atpA, atpB, atpF, ndhA, cemA, and DreamTaq polymerase (Thermo Scientific)—for genes petB, petD, ndhB, ndhF. Each cDNA-sample was analyzed with a lower (17–20) and a higher (22–25) number of cycles. Primer sequences are given in the Supplementary Material Table S1.

PCR products were analyzed by electrophoresis in a 1.5 % agarose gel with ethidium bromide staining, and the results were recorded on a Typhoon Trio+ scanner (GE Healthcare).

Electrophoresis and western blotting

Leaves were ground in ice-cold double Laemmli sample buffer (125 mM Tris–HCl, pH 6.8, 4.6 % SDS, 15 % glycerol, 10 % β-mercaptoethanol, 0.04 % bromophenol blue), boiled for 4 min, centrifuged for 1 min, and separated in 12 % SDS-PAGE according to (Laemmli 1970). For western blot analysis, proteins were transferred onto a nitrocellulose membrane (Schleicher&Schuell, Germany, 0.45 μm) as described previously (Towbin et al. 1979). After blotting, proteins were stained with chloramine T, KJ, and starch (Kumar et al. 1985). Membrane blocking and incubation with antisera were performed in buffer consisting of 20 mM Tris–HCl, pH 7.5, 150 mM NaCl, 0.05 % Tween 20, and 1 % ovalbumin at room temperature for 1 h. For protein analysis, we applied antisera from Agrisera: rabbit polyclonal antisera against chloroplast proteins Lhcb1 (AS01 004) and PsbO (AS06 142-33), and goat anti-rabbit IgG conjugated with horse radish peroxidase (AS09 602). After each incubation with antiserum, the membrane sheet was washed six times in PBS buffer with 0.05 % Tween 20. Visualization was performed using 4-chloro-1-naphthol as a substrate for the peroxidase.


The data were processed using Excel (Microsoft) software. Significance of the differences between mean values was verified by the Student’s t-test.


Growth inhibition

We have previously investigated the effect of cadmium on the growth of maize seedlings (Klaus et al. 2013). In the conditions used, cadmium at a concentration of 80 μM inhibited growth strongly, but this was far from a lethal effect, and all of the parameters studied differed from the control significantly after just 6 days of exposure. We used the same experimental design in the present research, with modifications to the light regime and observation of more parameters.

In the presence of 80 μM Cd, the growth of maize and barley seedlings was affected in a similar manner (Table 1, Supplementary Material Table S2). For many parameters, an inhibitory effect of approximately 30 % was found. The content of photosynthetic pigments was reduced by 10–20 % in barley and by 15–30 % in maize. The loss of water was significant. In both species, shoots were more depressed in the accumulation of masses, whereas roots lost more water. If the difference between the two species in relative strength of an inhibitory effect was significant, then the growth of barley seedlings was depressed more than that of maize (Table 1). The opposite was true for Chl content in first leaves and for roots (length and dry weight) only. In barley, some seedling parameters were reduced by half or more.
Table 1

Relative effect of 80 μM Cd on the growth of barley and maize seedlings (% of control plants)



H. vulgare

Z. mays



63.5 ± 0.5

72.3 ± 1.6**


1st leaf

77.7 ± 0.7

97.6 ± 1.4**

2nd leaf

69.5 ± 1.7

72.5 ± 2.0


71.4 ± 1.1**

61.4 ± 1.0


2nd leaf

51.6 ± 1.8

67.6 ± 2.4**

Fresh weight (FW)

1st leaf

72.0 ± 1.1

85.1 ± 1.8**

2nd leaf

50.2 ± 1.8

62.3 ± 2.1**


39.6 ± 0.8

64.3 ± 1.9**


55.7 ± 0.9

68.8 ± 1.8**


95.1 ± 2.0

93.4 ± 1.2


59.5 ± 1.5

73.1 ± 1.6**

Water content

1st leaf

98.9 ± 0.1

99.3 ± 0.2

2nd leaf

95.6 ± 0.2

98.9 ± 0.1**


96.4 ± 0.1

98.9 ± 0.1**


97.4 ± 0.2

99.1 ± 0.3**


94.9 ± 1.0

98.4 ± 0.1**


1st leaf

77.9 ± 3.6**

65.3 ± 1.1

2nd leaf

78.5 ± 2.7

80.4 ± 0.8


1st leaf

86.8 ± 4.0**

68.4 ± 1.7

2nd leaf

83.7 ± 4.4

81.6 ± 1.1


1st leaf

89.7 ± 4.9

79.6 ± 0.7

2nd leaf

88.9 ± 3.7

84.8 ± 1.2

Plant organs: leaf (1st or 2nd)—leaf blade only, stem—with leaf sheaths, shoot—stem with leaf sheaths and blades, caryopses—from which the seedling was grown, root—all roots of a seedling. All parameters of the Cd-treated seedlings are given as % of the same parameters in control seedlings (100 %) ± standard errors (SE)

** Difference between maize and barley seedlings in relative growth inhibition is significant, P ≥ 0.95 (asterisks at less inhibited species). Full version of the table with absolute values is in the Supplementary Material Table S2

Another specific response is the emergence of brown spots on leaves. On the fourth day in the presence of 80 μM Cd, the distal part of the barley leaves started to become covered with small brown spots (Fig. 1a). The spots were not induced by a 6-day exposure, nor were they induced with a lower cadmium concentration (e.g., 20 μM), nor with 100 μM of Cu or Ni, nor in the leaves of maize seedlings at any of these conditions. In seedlings grown in November, spots were hardly detectable at the end of the standard experiment. The chlorophyll fluorescence in barley leaves was reduced in these spots (Fig. 1b). Most likely, these spots represent necrosis stimulated by cadmium in the barley leaves.
Fig. 1

Spots on the first leaves of 9-day-old barley seedlings after a 6-day exposition to 80 μM Cd. a Visual image. b Chlorophyll fluorescence (Image-PAM data)

After 31 days of exposure to 250 μM Cd, 2.4 % of the barley plants died (N = 208). The lethal effect for maize was studied previously (Klaus et al. 2013).

Cadmium accumulation in organs and in chloroplasts

Seedlings of both species accumulated nearly the same quantity of Cd—40–45 μg (Table 2). Barley restrained more Cd in their roots. Caryopsis turned up additional depot for cadmium (they didn’t contact directly with cadmium-containing solution). Small barley caryopsis accumulated twofold more Cd than that of maize (29 times more per unit of weight). In 6 days, barley caryopsis gathered 8.9 % of cadmium adsorbed from mineral medium by whole seedling or 50 % of Cd penetrated “root barrier” (Table 2). Therefore, cadmium content in barley shoots is 2–3 times lower than in maize ones.
Table 2

Accumulation of Cd in organs of barley and maize seedlings


Cd, μg/g DW

Cd, μg per organ

H. vulgare

Z. mays

H. vulgare

Z. mays

Cd (μM)









1st leaf


234 ± 12c

1.9 ± 1.7a

626 ± 57c


2.5 ± 0.2c

0.01 ± 0.01a

4.0 ± 0.1a

2nd leaf


108 ± 13a

1.0 ± 1.0a

307 ± 51b


0.4 ± 0.1a

0.01 ± 0.01a

2.8 ± 0.3a

1st + 2nd leaves


199 ± 20b

1.3 ± 0.9a

439 ± 36b


2.8 ± 0.2c d

0.03 ± 0.02a

6.8 ± 0.4b


1.04 ± 0.01b

199 ± 9b

1.3 ± 0.7a

340 ± 33b

0.01 ± 0.00b

1.0 ± 0.1b

0.02 ± 0.01a

4.0 ± 0.3a



288 ± 37c


9.9 ± 2.4a


4.0 ± 0.5d


1.9 ± 0.3a


17.3 ± 6.6c

4642 ± 388d

9.9 ± 5.9a

2796 ± 199d

0.13 ± 0.05c

36.7 ± 2.9e

0.11 ± 0.06a

28.6 ± 1.8c

Whole seedling


0.14 ± 0.05

44.5 ± 3.1

0.16 ± 0.07

41.2 ± 1.7

Organ definitions as in Table 1. The data are given as mean ± SE. a–d Differences in Cd accumulation between organs of the same seedlings (P ≥ 0.95)

Then, we analyzed accumulation of Cd in chloroplasts isolated from the first two leaves and purified on a Percoll gradient. Cadmium in chloroplasts of control plants was below the level of detection (Table 3). At 80 μM Cd, barley chloroplasts contained 3.5 times more Cd than chloroplasts of maize (per mg Chl). We normalized the Cd content both in leaves and in chloroplasts to the Chl level, and we accepted the Cd content in leaves (μg Cd/mg Chl, Table 3) as 100 % and calculated the portion of cadmium in chloroplasts as X %. Such an approach allows us to show that the portion of net leaf Cd translocated into chloroplasts was low in barley (1.16 %) and very low in maize (0.21 %).
Table 3

Accumulation of Cd in chloroplasts


Cd in medium, μM

Cd in leaves, μg/mg Chl

Cd in chloroplasts

ng/mg Chl

% of Cd in leaves1

Cd cation:Chl molecules2

H. vulgare






14.8 ± 1.6b

171 ± 26b

1.16 ± 0.14b



9.5 ± 1.3a

126 ± 7ab

1.32 ± 0.09b


Z. mays


0.01 ± 0.01




23.6 ± 1.9c

49 ± 5a




26.7 ± 3.2c

92 ± 20ab

0.34 ± 0.03a


P. australis (Pietrini et al. 2003)3


0.95 ± 0.07

93 ± 7




2.35 ± 0.18

336 ± 24



B. napus (Baryla et al. 2001)3

100 mg/kg soil





The data are given as mean ± SE

n.d. not detectable, a–c differences significant at P ≥ 0.95

1Determined by equation: \(C/L \cdot 100\, \%\), where C is the Cd content in chloroplasts, L is the Cd content in leaves both μg/mg Chl

2Mr Chl was approximated as 900 (a = 893.5, b = 907.5)

3Recalculation of data from the article

4Variance cannot be calculated

5Water content is assumed to be 93 % (see text)

Additionally, we studied Cd penetration into chloroplasts at a lethal concentration of 250 μM Cd. Remember, that under the experimental design we used, dying of seedlings started later, than the experiment was completed. The Cd concentration tripled in nutrient medium, but this did not cause a similar rise of Cd in leaves (Table 3). At 250 μM Cd in maize leaves, the cadmium content per mg Chl was slightly higher compared to that at 80 μM Cd, but the content of Cd in chloroplasts (ng/mg Chl) increased nearly twofold, and the portion of chloroplastic Cd rose from 0.21 to 0.34 %. In barley leaves at lethal conditions, the Cd level declined significantly with a concomitant reduction of cadmium content in chloroplasts (ng/mg Chl); the portion of net leaf Cd that penetrated into chloroplasts did not change (1.32 %). Hence, at a lethal concentration of 250 μM Cd, the accumulation of the metal in chloroplasts of both species was very similar (92 ± 20 and 126 ± 7 ng/mg Chl)

To compare our data with previously obtained results, we completed Table 3 with recalculated values from the articles of Baryla and Pietrini with co-workers (Baryla et al. 2001; Pietrini et al. 2003). In the article by Baryla with coauthors, the data on Chl content are given per fresh weight (FW), the data on Cd content are given per dry weight (DW), and the data on relative water content are absent. The only sources of data found (unpublished) for water content were three PhD theses completed at our Institute of Plant Physiology. In these works, the water content of control soil-grown plants of two B. napus cultivars varied from 92 to 94 %, and once to 95 %. Therefore, we recalculated the ratios of Cd/Chl in the leaves of B. napus with the assumption that the water content was 93 %. Other data from these two articles were recalculated without any assumptions.

Effect on processes in thylakoid membranes

We used a wide range of concentrations, and the highest (200 μM Cd) is non-lethal (Klaus et al. 2013). In both species, cadmium did not affect significantly the photochemical parameters: potential (Fv/Fm) and effective (ΦPSII) quantum yield of PSII photochemistry, and photochemical quenching of the excited state of Chl (qP) (Fig. 2a, b). The most prominent effect was found in non-photochemical quenching of Chl excitation. In barley leaves, the parameter NPQ decreased exponentially as cadmium concentration increased in hydroponic medium (Fig. 2c). The decrease was caused by a decline in the fast relaxing component of NPQ (qE). In leaves of maize, neither NPQ nor its component qE were changed significantly (Fig. 2d). The slow relaxing component (qI) of NPQ and another coefficient of the non-photochemical energy quenching—qN—were not influenced by Cd, or they changed only slightly but in the opposite direction. In barley leaves, both qI of NPQ and qN were increased significantly at 80 μM Cd, and qI—at 20 μM Cd. The same tendency—rise of qI and fall of qE—was seen in the second species at the two highest Cd concentrations. However, in maize, these changes were minor, and the only significant difference from the control was an increase in qI at the highest concentration studied.
Fig. 2

Changes in chlorophyll fluorescence parameters induced by 4, 20, 80, and 200 μM Cd in leaves of barley and maize seedlings a, bFv/Fm, ΦPSII, qP, qN; c, d NPQ and its fast (qE) and slow (qI) relaxing components; a, cH. vulgare; b, dZ. mays. Differences from control, significant at *P ≥ 0.95; **P ≥ 0.99

To verify the specificity of cadmium action on processes in thylakoid membranes, we studied the effect of another HM. We chose copper as an element similar in its toxic action and different in its chemical nature. Both metals were applied at high concentration—100 μM. In barley seedlings, both HMs affected processes in the thylakoid membranes similarly: photochemical parameters (Fv/Fm, ΦPSII, qP) were not changed, and NPQ was reduced due to a decrease in qE, but qI was increased to a lesser extent (Table 4). Both cations changed these parameters in barley to nearly the same values. The only difference was found in their impact on qN. Copper did not affect the coefficient, but 100 μM Cd increased qN by 20 %. In the previous experiment (Fig. 2a), 80 μM Cd increased qN to the same extent, but the deviation of the data was smaller and the difference from the control was significant (P ≥ 0.99).
Table 4

Changes in chlorophyll fluorescence parameters induced by 100 μM of Cd or Cu in leaves of barley and maize seedlings







NPQ fast (qE)

NPQ slow (qI)

H. vulgare


0.76 ± 0.02

0.63 ± 0.04

0.92 ± 0.05

0.31 ± 0.01

1.20 ± 0.01

0.86 ± 0.02

0.34 ± 0.01


0.79 ± 0.02

0.66 ± 0.02

0.95 ± 0.01

0.37 ± 0.04

0.68 ± 0.04**

0.26 ± 0.05**

0.42 ± 0.02**


0.80 ± 0.02

0.67 ± 0.04

0.92 ± 0.03

0.31 ± 0.02

0.67 ± 0.03**

0.25 ± 0.04**

0.42 ± 0.02*

Z. mays


0.79 ± 0.03

0.54 ± 0.03

0.75 ± 0.02

0.29 ± 0.02

0.69 ± 0.01

0.49 ± 0.03

0.20 ± 0.01


0.79 ± 0.03

0.57 ± 0.03

0.80 ± 0.02

0.30 ± 0.01

0.69 ± 0.02

0.46 ± 0.03

0.23 ± 0.03


0.78 ± 0.03

0.62 ± 0.03

0.87 ± 0.02**

0.31 ± 0.02

0.48 ± 0.02**

0.19 ± 0.03**

0.29 ± 0.02**

The data are given as mean ± SE. Difference from control significant at: * P ≥ 0.95; ** P ≥ 0.99

In maize, the copper effect was similar: quantum yield (Fv/Fm, ΦPSII) and qN were not influenced, NPQ and qE were reduced, and qI increased. Specifically, Cu stimulated photochemical quenching (qP) by 17 % (Table 4). None of the parameters studied was changed in maize by treatment with 100 μM Cd (Table 4). These data imply that the absence of a qE decline is specific to maize seedlings at non-lethal concentrations of cadmium.

mRNA level in chloroplasts

To reveal potential sources of qE decline, we investigated the effect of Cd on the accumulation of mRNAs in chloroplasts. We chose 9 plastome genes that code for polypeptides of the thylakoid complexes that affect the proton gradient: cytochrome b6/f (petB, petD, petL), ATP synthase (atpA, atpB, atpF), and NADH dehydrogenase (ndhA, ndhB, ndhF)—and one extra gene that encodes a protein involved in transport across the plastid envelope membrane (cemA) (Rolland et al. 1997). We analyzed the mRNA levels of these genes in chloroplasts from control and cadmium-treated seedlings. Maize and barley seedlings grown at 80 μM Cd showed no changes in mRNA levels (Fig. 3). Therefore, we tested the effect of 250 μM Cd, which in a prolonged experiment is lethal for both maize (Klaus et al. 2013) and barley (see above). However, no changes in the steady-state levels of the 10 chloroplast mRNAs occurred, even under such extreme conditions (Fig. 3).
Fig. 3

Relative mRNA levels in chloroplasts of seedlings grown at 0 (control), 80, or 250 μM Cd. Semi-quantitative RT-PCR analysis

Protein level in chloroplasts

Finally, we examined Cd impact on the levels of chloroplast proteins. The general pattern of proteins was similar in leaves of both control seedlings and of seedlings grown at 80 and 250 μM Cd (Fig. 4a). Bands with apparent molecular masses of approximately 55 kDa—huge in the C3-plant barley and moderate in the C4-plant maize—most likely represent the large subunit (LS) of Rubisco. In both species, the band in control seedlings was substantially bigger than in seedlings grown at 250 μM Cd. In 80 μM Cd-treated seedlings, the relative intensity of the band varied: sometimes it resembled the control, or the 250 μM Cd-treated seedlings, or displayed an intermediate intensity.
Fig. 4

Effect of cadmium on protein levels in leaves of seedlings grown at 0 (control), 80, or 250 μM Cd. a Staining for total leaf protein. Asterisk with arrowpoint to plausible position of LS Rubisco. b Immunoblot analysis of Lhcb1 and PsbO proteins

We applied antisera against two thylakoidal proteins—antenna protein Lhcb1 and large polypeptide from the PSII oxygen-evolving complex OEC33/PsbO. Cadmium decreased the level of both proteins (Fig. 4b). The decrease in maize leaves was more obvious and gradual than in barley leaves.


We analyzed cadmium impact on two Poaceae species—barley and maize. The two species responded similarly in many respects. Whole seedlings accumulated 40–45 μg Cd (Table 2), and their growth was inhibited rather similarly (Table 1). Photochemical parameters—Fv/Fm, ΦPSII, qP (Fig. 2a, b)—and the levels of 10 mRNAs in chloroplasts (Fig. 3) were not changed. The levels of some chloroplast proteins were reduced in cadmium-treated seedlings of both species (Fig. 4).

However, many reactions were species-specific. Cadmium (80 μM and higher) induced the appearance of brown spots on leaves of barley seedlings (Fig. 1a). The effect was very specific. The dots were not induced by Cu or Ni, and they were not induced in the leaves of maize seedlings. The dots were hardly detectable in late autumn. Maximum chlorophyll fluorescence (Fm) was reduced greatly in these spots (Fig. 1b). The induction of brown dots by cadmium was described in leaves of woody species previously (Pietrini et al. 2010). In poplar, the effect was cultivar-specific. Brown spots accumulated much more cadmium than the adjacent green tissues. Authors suggested that these spots are necrotic areas where cadmium can be excluded (Pietrini et al. 2010).

Barley and maize accumulated different amounts of Cd in their chloroplasts (Table 3). However, the difference between previously available data was enormous. Chloroplasts of B. napus contained 4.5 ng Cd/mg Chl, whereas chloroplasts of P. australis contained 93 and 336 ng Cd/mg Chl. The percentage of Cd in chloroplasts compared to Cd in leaves differed even more drastically: 0.02 % in rape and 9.8 and 14.3 % in common reed (original data recalculated in Table 3). Our data filled the gap, and we can see relatively continuous rows for both the Cd content in chloroplasts (ng Cd/mg Chl)—4.5–49–92–93–126–171–336—and the percentage of leaf cadmium that penetrated into chloroplasts: 0.02–0.21–0.34–1.16–1.32–9.77–14.29 % (Table 3). In Pietrini’s experimental model, plants accumulated extremely low levels of Cd in their leaves, which in turn made the portion of chloroplastic Cd incredibly high (10–15 %). In that work, plants were grown from collected rhizomes, not from seeds (Pietrini et al. 2003). The only research performed with soil-grown plants (Baryla et al. 2001) demonstrated very low level of Cd penetration into chloroplasts (Table 3). Probably, most of Cd in a soil is bound, and its concentration in free soil solution hardly exceeds 1 μM. Thus, we suppose that in natural conditions, the values of Cd accumulation in chloroplasts will be closer to the data obtained by Baryla and co-workers.

The data from Table 3 show that there is no clear correlation between the concentrations of Cd in nutrient medium in its penetration into chloroplasts. At 80 μM Cd and in the same experimental conditions, barley chloroplasts accumulated several times more Cd than maize chloroplasts. Three Poaceae species had different reactions to a rise in Cd concentration. In P. australis, increase from 50 to 100 μM Cd was followed by a growth in Cd content in both leaves and chloroplasts, and in the percentage of chloroplastic Cd. In Z. mays, an increase from 80 to 250 μM Cd barely affected the Cd content in the leaves, but the Cd content in chloroplasts and its proportion in the leaf pool rose. Both species belong to the PACMAD clade. H. vulgare is of other group—the BEP clade—and its response is very different. A rise from 80 to 250 μM Cd decreased the cadmium accumulation in barley leaves by one third, with a concomitant fall in Cd content in the chloroplasts (ng/mg Chl); the portion of the leaf Cd that invaded the chloroplasts remained unchanged. Therefore, we need more information to determine whether increasing the cadmium concentration in a nutrient solution will be followed by a concomitant increase in Cd content in the chloroplasts.

There is no doubt that in vitro Cd2+ can substitute Mg2+ in the porphyrin ring of Chl, and the characteristics of such modified Chl will be changed substantially. Some researchers cited an earlier article (Küpper et al. 1998) to support the opinion that this chemical reaction may underlie the mechanism of Cd toxicity on the photosynthetic apparatus and plant growth. However, if plants were grown naturally and cadmium uptake was performed by the roots, then the Cd:Chl ratio in the chloroplasts was very low (Table 3). Typically, one Cd cation per thousand(s) Chl molecules was recorded; the maximal ratio was one Cd per 370 Chl. In soil-grown plants, the ratio dropped to one Cd per more than 25,000 Chl (Table 3). If detached leaves were placed in a solution with an enormous concentration of 5 mM Cd and cadmium translocation was not restricted by the “root barrier,” then the Cd:Chl ratio in fragments of the PSII-enriched thylakoid membranes was also “one-to-thousand(s)”; the maximal ratio reached 1:257 (Geiken et al. 1998). If we assume that none of the Cd particles in the chloroplast are complexed with glutathione (of the ascorbate–glutathione cycle) or bound by thiol groups of proteins, then less than 1 % of the antennae Chl can be attacked by Cd2+. Therefore, the hypothesis that substitution of Mg2+ by Cd2+ in antennae Chl can substantially influence light absorption by leaves and impact plant growth should be rejected.

In barley and maize seedlings, cadmium did not significantly alter the photochemical parameters (Fv/Fm, ΦPSII, qP) (Fig. 2a, b). Two non-photochemical coefficients—qN and NPQ—were changed independently. Such a discrepancy was described previously, and it is known that qN and NPQ can be changed even in an opposite direction (Lichtenthaler et al. 2005). Both coefficients operate with Fm and Fm values, but calculation of qN is based on the F0 value and NPQ is not.

NPQ was almost unchanged in chloroplasts of maize, but NPQ in barley chloroplasts decreased substantially (Fig. 2c, d). The reactions of plant photochemistry to cadmium are diverse (see “Introduction” section), but Cd changes non-photochemical quenching of the excited state of Chl (if at all) in only one direction—up (Pietrini et al. 2003, 2010; Filek et al. 2010; Cai et al. 2011). Still we found one article showing that Cd reduces NPQ, but in that study cadmium was applied to leaf disks, not to intact plants (Liu et al. 2010). We have conducted perhaps the first study to identify the component that causes the change in NPQ in the presence of Cd. The concentration-dependent decrease in NPQ was dependent on the fast relaxing component qE (Fig. 2c). qE depends on the trans-thylakoidal proton gradient, and it characterizes the energization of the thylakoid membranes. The slow relaxing component of NPQ—qI—was not changed or was only slightly increased. The parameter depends on the translocation of the light-harvesting complex apart from PSII (state transition from “state 1” to “state 2”) and on the photoinhibition of PSII. The decrease in qE is dependent on external Cd concentration and is species-specific, whereas the increase in qI can be coupled with Cd accumulation in chloroplasts in both species. The highest Cd content in maize chloroplasts was found at 250 μM Cd (Table 3), and the greatest increase in qI (32 %, P ≥ 0.95) was seen at 200 μM Cd (Fig. 2d). In barley, the highest Cd content in chloroplasts (Table 3) and the most significant rise in qI (23 %, P ≥ 0.99, Fig. 2c) were observed at 80 μM Cd; the only significant increase in qN (19 %, P ≥ 0.95) was also observed at this concentration (Fig. 2a). At higher Cd external concentrations (200/250 μM), the values of qI (Fig. 2c), qN (Fig. 2a), and Cd content in chloroplasts (Table 3) were smaller compared to those at 80 μM Cd.

These data suggest that Cd in chloroplasts (1) depletes pH gradient across the thylakoid membrane and (2) stimulates photoinhibition and/or state transition of the PSII antennae. The first effect, probably, is more sensitive to Cd concentration in a nutrient medium, whereas the second effect is more sensitive to Cd concentration inside the chloroplasts. The first effect decreases non-photochemical quenching, and the second effect increases it. If these changes are equal in absolute value, then the net effect may be around zero. That last speculation may be illustrated by Fig. 2d. At 200 μM Cd (and, probably, to a lesser extent at 80 μM Cd), qI is increased, qE is decreased, and the net NPQ is close to the control value. An additional example comes from the changes of qN in leaves of pea and broad bean: qN is increased at 50 and 250 μM Cd (the second effect prevails), it remains close to the control level at 1 mM Cd (effects are equal), and qN goes down at 5 mM Cd (the first effect is predominant) (Geiken et al. 1998). Hence, in this study the first effect also increased along with Cd concentration in a nutrient medium.

There are two ways to diminish ΔpH: downregulate its generation or upregulate its utilization. We studied chloroplast mRNAs which code for subunits of complexes in thylakoid membranes that can be involved in the process: cytochrome b6/f (petB, petD, petL), ATP synthase (atpA, atpB, atpF), NADH dehydrogenase (ndhA, ndhB, ndhF), and one extra gene that encodes protein involved in transport across plastid envelope membrane (cemA) (Rolland et al. 1997). In both species, expression of these genes was not changed at the level of mRNA even at lethal concentration 250 μM Cd (Fig. 3). We cannot nevertheless exclude their alteration at the protein level or at the level of functional activity. We can suppose that Cd lowered generation of ΔpH. It is known that Cd inhibits PSII (Bazzaz and Govindjee 1974; Geiken et al. 1998) and oxygen evolution (Faller et al. 2005; Pagliano et al. 2006; Rascio et al. 2008). However, in our experiments quantum yield of PSII did not change (Fig. 2). The level of the large subunit of water-oxidizing complex (PsbO) was downregulated, but changes in qE and PsbO level were not correlated. In barley leaves PsbO level decreased slightly (Fig. 4), and qE decreased greatly (Fig. 2). In maize leaves, vice versa, PsbO level decreased greatly (Fig. 4), but qE level did not change (Fig. 2). Hence, this assumption cannot be supported by our results.

Alternative explanation is that Cd stimulated utilization of ΔpH. Under continues stress, plants may activate their metabolism and ATP consumption which upregulates ATP synthase activity and may decrease ΔpH. Some researchers referred to such a possibility (e.g., Stirbet et al. 2014; Kalaji et al. 2014). We discussed above that qE changes in barley correlate better with external Cd concentration than with Cd accumulation in chloroplasts. This implies indirect action through alteration of some metabolic processes in the whole organism. Stability of qE in maize can also be explained by indirect action. Cd and Cu inhibit enzyme of C4 metabolic pathway—phosphoenolpyruvate carboxylase (PEPC) (Iglesias and Andreo 1984). In vitro, Cu effect is bigger and can be reversed by a thiol agent, Cd effect is smaller but irreversible. Probably, in vivo, maize can alleviate Cu toxic effect (e.g., by glutathione in chloroplasts) and maintain metabolism, ATP consumption, and synthesis at high level, which lead to decrease in qE level (Table 4). Perhaps, maize is unable to protect PEPC from Cd action, and therefore metabolism, ATP usage, and synthesis are low, and qE is unchanged (Table 4). Barley has C3 type of photosynthesis, and both Cd and Cu lowered qE (Table 4), apparently through the mechanism discussed. Inactivation of PEPC by Cd may be responsible for larger decrease in levels of Chl, Chl-binding protein (Lhcb1), and PsbO in maize (C4) as compared to barley (C3) (Table 1; Fig. 4). Therefore, it looks more probable that ΔpH was declined in barley due to metabolic activation of ATP synthesis, whereas in maize this process was abolished through inhibition of PEPC and C4 pathway.

The levels of chloroplast proteins are usually unchanged in Cd-treated plants. This was shown for both dicotyledonous (Franco et al. 1999; Baryla et al. 2001; Liu et al. 2008) and monocotyledonous (Pagliano et al. 2006; Rascio et al. 2008; Janik et al. 2010) species. The Cd impact was studied for a handful of chloroplast proteins: D1, D2, Lhcb (1, 2, 3), LS, and small (SS) subunits of Rubisco, and probably PsbO (OEE1 in (Pagliano et al. 2006)). Content of these proteins was mainly determined in thylakoids, but the levels of LS and SS Rubisco were not changed in whole leaves of tomato plants (Liu et al. 2008). In rice, the levels of proteins D2, Lhcb, and OEE1 (PsbO) remained unchanged at 25 and 75 μM Cd, and bands of D1 in Cd-treated plants were weaker than in control plants. However, the authors considered the variation as insignificant (Pagliano et al. 2006). In another article, the level of D1 in rice was not changed at 50–250 μM Cd (Rascio et al. 2008). However, in leaves of P. australis, the level of LS Rubisco decreased significantly (Pietrini et al. 2003). All of these works dealt with C3-plants. Our data obtained from the C3-plant H. vulgare are in good agreement with them: the levels of chloroplast proteins Lhcb1, PsbO, and LS Rubisco were reduced slightly (Fig. 4). The reductions of LS Rubisco in P. australis (Pietrini et al. 2003) and H. vulgare (Fig. 4a) are quite similar. We failed to find any articles that studied Cd impact on chloroplast proteins in C4-species. In our experiment, levels of the chloroplast proteins Lhcb1, PsbO, and LS Rubisco in the leaves of the C4-plant Z. mays were greatly reduced (Fig. 4). According to the immunoblotting data, the reductions of Lhcb1 and PsbO levels were concentration-dependent: at 250 μM Cd the effect was more pronounced than at 80 μM Cd. These data imply that levels of chloroplast proteins in C4-plants may be more sensitive to cadmium action than in C3-species. For example, due to interruption of C4 pathway through PEPC inhibition (Iglesias and Andreo 1984). However, such a conclusion must be verified by analyzing more species and proteins.


The authors thank Dr. N. Pedentchouk (University of East Anglia) for kindly providing the antiserum to the Lhcb1 protein. This work was supported by the grant from the Russian Science Foundation (No. 14-14-00584).

Supplementary material

11120_2014_47_MOESM1_ESM.pdf (106 kb)
Supplementary material 1 (PDF 105 kb)

Copyright information

© Springer Science+Business Media Dordrecht 2014

Authors and Affiliations

  • Eugene A. Lysenko
    • 1
  • Alexander A. Klaus
    • 1
  • Natallia L. Pshybytko
    • 2
  • Victor V. Kusnetsov
    • 1
  1. 1.Timiryazev Institute of Plant Physiology, RASMoscowRussia
  2. 2.Institute of Biophysics and Cell Engineering, NASBMinskBelarus

Personalised recommendations