Growth and seedling development
Each MFC contained eight seed cassettes with ten seeds each. Seed germination in the two MFCs with HGMFs was higher than in the non-magnetic chamber (Table 1), indicating that the presence of magnets and magnetic fields did not negatively affect germination.
Table 1
Germination and root length of Brassica rapa seeds in magnetic field chambers with (A&B) and without magnets (C). There was no difference between chambers or seed cassettes.
|
|
|
Cassette:
|
|
|
|
1
|
2
|
3
|
4
|
5
|
6
|
7
|
8
|
MFC-A
(HGMF)
|
Average:
|
14.3
|
14.6
|
15.4
|
11.4
|
14.4
|
15.1
|
16.2
|
12.6
|
14.7
|
SE:
|
4.2
|
3.2
|
3.5
|
3.7
|
4.4
|
2.8
|
4.8
|
4.8
|
5.0
|
N=
|
77
|
9
|
10
|
10
|
10
|
10
|
10
|
9
|
9
|
MFC-B
(HGMF)
|
Average:
|
16.1
|
16.6
|
15.0
|
17.3
|
16.8
|
15.5
|
16.3
|
16.4
|
14.6
|
SE:
|
3.5
|
2.5
|
3.0
|
3.6
|
4.9
|
3.5
|
3.7
|
4.0
|
2.8
|
N=
|
80
|
10
|
10
|
10
|
10
|
10
|
10
|
10
|
10
|
MFC-C:
(no magnets)
|
Average:
|
14.7
|
19.1
|
14.2
|
13.4
|
15.2
|
15.3
|
14.5
|
13.0
|
13.4
|
SE:
|
4.4
|
3.9
|
4.7
|
4.0
|
2.3
|
4.3
|
6.4
|
3.3
|
3.2
|
N=
|
70
|
8
|
10
|
9
|
8
|
9
|
10
|
7
|
9
|
Curvature Induction By Hgmf:
Exposure to HGMF induced curvature (Sup. Video 1, Table 2). However, substantial curvature occurred in both presence and absence of magnetic gradients. The non-magnetic chamber showed 81% curvature in roots within the seed cassette around the germination paper (Fig. 2).
Table 2
Germination and curvature in the presence and absence of HGMFs of a total of 80 seeds per MFC
|
not germinated
|
Straight*
|
curved
|
curved (HGMF)
|
MFC-A
(HGMF)
|
3
|
8
|
40
|
28
|
MFC-B
(HGMF)
|
0
|
0
|
50
|
30
|
MFC-C
(No magnets)
|
10
|
5
|
65
|
0
|
*refers to roots that never emerged from the seed cassette or were too short to reach the HGMF area (as shown in Fig. 2). |
MFCs with magnets showed curvature that was independent of the HGMF because of growth patterns or curvature occurring inside the seed cassettes where the magnetic gradient was too weak to affect curvature.
The comparison of space-grown or static ground control seedlings exposed to HGMF with seedlings grown in the non-magnetic field chamber (#1 and #23 in Sup. Table 2) shows strong correspondence between the two conditions (Fig. 3), indicating that magnetic fields do not affect general metabolism or transcription activity. This notion is in line with reports that failed to detect effects of magnetic fields on growth50,51 but contrasts with effects of weak magnetic fields on root curvature52 and effects of the geomagnetic field on stress response and hormesis53,54. The geomagnetic field (typically about 0.5 G) is orders of magnitude weaker than the employed magnetic fields in this research (ca. 30 kG).
These data indicate that factors other than HGMF induce curvature and the most likely factor is hydrotropism18,30,55. The lack of a gravity stimulus and the distance between the germination paper and the HGMF-generating wedges resulted in a large number of roots confined to the germination paper. The same factor also applied to seedlings in the magnetic chambers and resulted in reduced numbers of roots reaching the HGMF. However, the unambiguous evidence for curvature in space (see Sup. Video 1) and clinorotated seedlings (Sup. Video 2) indicates that HGMFs do have the ability to induce curvature as had been demonstrated earlier in roots, coleoptiles, inflorescences, and hypocotyls16,17,24,56,57.
The presumably hydrotropic growth reduced the number of roots that reached the influence of the HGMFs. Therefore, the overall number of roots curving in response to the magnetic gradient was low (Table 2).
Examining the gravisensing mechanism includes measuring the size of the presumptive sensors as previous work has shown that gravisensitivity depends on the amount of starch in amyloplasts58. Assuming equal density, the diameter of amyloplasts determines the relative mass. Measurements of amyloplasts in columella cells differed between the space-grown, clinorotated, and statically grown ground controls (Fig. 4) and indicates that the amyloplast size and therefore the gravisensing mechanism is responsive to the growth condition, notably weightlessness and clinorotation that respectively reduces and enhances mechanostimulation.
Earlier analyses of starch from space and ground controls found that ethylene reduces starch particle size and that in cotyledons starch particles sizes in space and ground controls were of equal size59. Since we used activated charcoal cloth to absorb volatile organic compounds (Fig. 1), we assume that ethylene was not a factor in the different experiments. In addition, we examined the size distribution of amyloplasts in the root columella. Earlier work showed that amyloplasts in gravisensing tissues (root cap and endodermis) are about twice the size of other tissues7. Our data (Fig. 4) indicate that the amyloplast size is responsive to the gravitational and mechanical stimulation. The reduced size after clinorotation, the average size in static 1-g controls, and the enlarged size in roots grown under ‘micro-gravity’ conditions support the notion that the extend of gravitational stimulation is inversely proportional to the mass of the amyloplasts. Thus, plants not only perceive the direction of an accelerative force but adjust their (starch) metabolism according to the amount of stimulation. If this notion is correct, then gene transcription data should support this notion and the following section confirms that amylase activity indeed is a function of mechanostimulation.
Transcript Analyses
Since four different tissue types were analyzed, tissue variability and response to spaceflight and clinorotation can be assessed for all examined tissues and genes. Based on distributions of transcription patterns (Figs. 3 and 5), a comprehensive analysis of all treatment and tissue combinations was performed such that the scatter for each comparison and gene was determined based on the formula\(\sum _{i=1}^{n}\sum _{j=1}^{4}\sqrt{{\left({X}_{ij}-\stackrel{-}{{X}_{j}}\right)}^{2}+{\left({Y}_{ij}-\stackrel{-}{{Y}_{j}}\right)}^{2}}\)
where j represents the tissue types (root tip, root proper, hypocotyl, cotyledons) and i the individual experimental comparisons (28, Sup. Table 2) or examined genes (16). This value was calculated for each analyzed gene and the smallest value (scatter in Figs. 3 & 5) was identified as the most stably transcribed gene (Table 3).
The distribution of the least variable gene transcription varied greatly (Table 3). Although common reference genes (e.g., TUB1, ACT7) are represented (Table 3), the observation that individual tissues differed in the genes of greatest stability and that the average of all tissues resulted in different assortments suggested that referring transcription data to the average of all measurements (i.e., the regression lines in Figs. 3 and 5) is superior to relying on a single gene. Because PFK showed the greatest stability for all tissues, we placed PFK at the origin of the coordinate system (0/0). However, the results explained below are independent of this selection.
Table 3
Assessment of the consistency of transcription data based on the average of examined tissues. The top four choices of all combinations (as in Sup. Table 2) are based on the percentage that resulted in the least scatter of all 448 combinations (16 genes by 28 comparisons).
|
Gene
|
n
|
%
|
All tissues
|
PFK
|
60
|
13.4
|
PIN1
|
47
|
10.5
|
TUB1
|
45
|
10.0
|
GLK
|
33
|
7.4
|
Root tip
|
PIN1
|
43
|
9.6
|
PIN3
|
41
|
9.2
|
SUS
|
40
|
8.9
|
GLK
|
35
|
7.8
|
Root
proper
|
COX
|
51
|
11.4
|
ACT7
|
42
|
9.4
|
SUS
|
42
|
9.4
|
HXK
|
39
|
8.7
|
Hypocotyl
|
HXK
|
42
|
9.4
|
ACT7
|
37
|
8.3
|
SUS
|
37
|
8.3
|
ADH1
|
27
|
6.0
|
Cotyledons
|
PFK
|
49
|
10.9
|
ACT7
|
45
|
10.0
|
ADH1
|
40
|
8.9
|
PIN7
|
35
|
7.8
|
The comparisons (Sup. Table 2) show the least and most significant effect on gene expression and can be used to identify the conditions that induce physiological responses in brassica seedlings. The greatest stability in gene transcription was seen for treatments with similar mechanical load, for example, static growth with and without HGMF, or clinorotated samples (KSC and Lab). The largest scatter or least consistent transcription pattern was associated with different mechanical loads such as flight (i.e., no mechanical load) and clinorotation (enhanced mechanical load), or static and space flight conditions. The main conclusion of these evaluations is that HGMFs or more generally, strong magnetic fields, do not affect transcription; HGMF data are equally present in the most and least affected conditions (Table 4). However, a comparison between clinorotated experiments with the original flight hardware at KSC and experiments in our lab (comparison # 17 and # 22) show statistically significant differences (Sup. Figure 4).
Sensitivity of AMY1
Data sets comparing space-grown with clinorotated seedlings show large scatter and offset (Fig. 5A and Sup. Figure 3A), suggesting that the shift in the transcription pattern is not dependent on HGMF but differences in mechanostimulation (i.e., clinorotation). This effect is especially noticeable for AMY1. The effect of clinorotation on amylase transcription is independent of the HGMF because transcription of AMY1 in the absence of HGMF was higher in static seedlings than clinorotated seedlings (Fig. 5B). Space flight and clinorotation resulted in reduced and elevated transcription, respectively. A similar pattern was observed between clinorotated and static samples; however, HGMF had no effect (Sup. Figure 3). The altered AMY1 transcription between clinorotation at KSC and our lab is related to different stabilities of the hardware. The flight hardware contained a webbing-like base mount (implemented because of weight concerns), which provided stable support during space flight but flexed readily during clinorotation compared to a rigid assembly on the laboratory clinostat.
The transcription data correspond with the observed size distribution of amyloplasts (Fig. 4) and indicate that amyloplast size is regulated by starch degradation in clinorotated seedlings and amylase repression (starch accumulation) in space samples. Together these observations indicate that plant adapt to the weightlessness of space by increasing their amyloplast size which likely enhances their gravisensitivity. In contrast, clinorotation represents excessive mechanostimulation and leads to a reduction of amyloplast size through enhanced degradation (amylase transcription).
The observed AMY1 levels between space-grown and clinorotated seedlings were independent of reference genes; the effect persisted regardless of whatever gene was used as reference. This observation supports using transcription of all available genes as a reliable approach to identifying transcriptional changes of individual genes.
The AMY1 data correspond to earlier observations of enhanced gravisensitivity of space-grown lentil seedlings60,61 and reduced starch after clinorotation62. Although the STS-107 flight experiment could not be retrieved to measure amyloplast size, the image analysis of flax seedlings indicated that the magnetic gradient had stronger effects than during previous ground experiments because the root curvature started at a greater distance from the HGMF-inducing wedge23 than during ground controls. This observation is in line with the present data and strongly supports greater (gravi)sensitivity of plants growing in a microgravity environment.
The current report is the first to associate starch metabolism with amyloplasts size and gravisensitivity. The larger amyloplasts in space-grown plants suggest that the application of HGMFs in space is more effective than in earth-grown and especially in clinorotated plants. However, the unreliable growth direction of roots makes HGMF difficult to implement.
Gravisensitivity has also been linked to changes in calcium in statocytes 63–65 and calcium has been shown to stabilize α-amylase66–68. Therefore, Ca2+ and amylase are controlling element for the starch content in statocytes. However, starch content is the result of homeostasis for catabolic and anabolic metabolism. Starch biosynthesis depends on a complex set of enzymes that include phosphoglucose isomerase (PGI)69, phosphoglucomutase (PGM)70,71, and starch synthases (SSs)69,72 among others. Data on the balance between starch degradation and starch biosynthesis undoubtedly would provide a more comprehensive assessment of the sensitivity of starch metabolism to reduced and enhanced mechanostimulation. However, the limited set of transcriptionally analyzed genes prevents a more thorough assessment of starch biosynthesis in response to altered mechanostimulation. Future space experiments might remedy this shortcoming.
Table 4
The similarity of gene transcription between pairs of treatments based on high gradient magnetic fields (HGMF) or no magnetic fields (no-MF) that were applied during the space flight (Flight), on ground controls (static), or during clinorotation (Clino) at the Kennedy Space Center (KSC) or in the laboratory (Lab). The values were obtained by averaging the R2 values that resulted from using each of the 16 genes as reference (i.e., origin in graphical comparisons such as Figs. 3&5 and Sup. Figure 3&4).
|
#*
|
R2, %
|
STDEV,%
|
Comparisons (all genes)
|
least stable <────────────────────────────────> most stable
|
23
|
94.8
|
1.9
|
KSC static no-MF vs. KSC static HGMF
|
17
|
90.8
|
2.7
|
Lab Clino HGMF vs. KSC Clino HGMF
|
28
|
82.6
|
7.5
|
Lab Clino no-MF vs. Lab Clino HGMF
|
1
|
80.2
|
7.2
|
Flight-no MF vs. Flight-HGMF
|
15
|
80.0
|
3.3
|
KSC static HGMF vs. KSC Clino HGMF
|
18
|
78.8
|
3.7
|
Lab Clino no-MF vs. KSC Clino HGMF
|
16
|
76.6
|
3.3
|
KSC static no-MF vs. KSC Clino HGMF
|
25
|
76.3
|
2.6
|
Lab Clino no-MF vs. KSC static HGMF
|
27
|
75.6
|
4.7
|
Lab Clino no-MF vs. KSC static no-MF
|
24
|
73.1
|
5.2
|
Lab Clino HGMF vs. KSC static HGMF
|
19
|
72.8
|
7.0
|
KSC static HGMF vs. KSC Clino no-MF
|
26
|
70.7
|
5.5
|
Lab Clino HGMF vs. KSC static no-MF
|
20
|
70.4
|
6.7
|
KSC static no-MF vs. KSC Clino no-MF
|
14
|
60.6
|
7.7
|
KSC Clino no-MF vs. KSC Clino HGMF
|
13
|
59.2
|
11.1
|
Lab Clino no-MF vs. Flight-no MF
|
22
|
56.9
|
6.2
|
Lab Clino no-MF vs. KSC Clino no-MF
|
21
|
56.2
|
7.3
|
Lab Clino HGMF vs. KSC Clino no-MF
|
7
|
56.1
|
11.2
|
Lab Clino no-MF vs. Flight-HGMF
|
12
|
53.6
|
7.0
|
Lab Clino HGMF vs. Flight-no MF
|
9
|
52.3
|
10.0
|
KSC Clino no-MF vs. Flight-no MF
|
10
|
51.0
|
8.2
|
KSC static HGMF vs. Flight-no MF
|
11
|
50.7
|
8.4
|
KSC static no-MF vs. Flight-no MF
|
6
|
49.9
|
7.5
|
Lab Clino HGMF vs. Flight-HGMF
|
8
|
46.0
|
7.1
|
KSC Clino HGMF vs. Flight-no MF
|
2
|
40.6
|
6.7
|
KSC Clino HGMF vs. Flight-HGMF
|
5
|
38.0
|
7.4
|
KSC static no-MF vs. Flight-HGMF
|
3
|
37.4
|
10.6
|
KSC Clino no-MF vs. Flight-HGMF
|
4
|
37.4
|
10.6
|
KSC static HGMF vs. Flight-HGMF
|
* Numbers refer to Sup. Table 2. |
Genes other than AMY1
Although weightlessness is the dominant difference between ground and space flights, the lack of density-driven gas exchange and water distribution are equally significant alterations for plant growth in space. Because the flight hardware was enclosed in a hermetically sealed (triple-contained) chamber, atmospheric effects can be excluded. Therefore, the following considerations only apply to gravity effects. Comparing the number of significantly (|p| < 0.05) affected combinations (genes by reference), shows the largest effect on AMY1 transcription (Fig. 6).
Genes other than AMY1 responded to specific conditions but at reduced power. ACT7 showed modifications only when comparing clinorotation at KSC with static growth (Comparisons #19 & #20); ACT7 was not affected by HGMF but sensitive to vibrations (higher frequency movements in excess to the 1.5 rpm of the clinostat). This relationship is similar to the enhanced AMY1 transcription addressed in Sup. Figure 4. The enhanced oscillation can be verified by the jitter in Sup. Video 2. To improve visibility, the individual frames were aligned, but the inconsistent position of imprinted data confirms the added vibrational stimulation that affected ACT7 activation (and enhanced AMY1 transcription) compared to regular (smooth) clinorotation.
COX showed the strongest upregulation in space flight material compared with static growth conditions (#4, #5, #10 #11) but was not affected by HGMF. This observation further supports the claim that magnetic fields do not affect transcription. Instead, COX responds to space flight associated stress as has been shown previously for fish brain73 and skeletal muscle74. Changes in G6PDH5 were limited to treatment differences between space flight and clinorotation (#8, #12, #15, #16) and correspond to earlier reports of altered enzyme activities of pine seedlings after exposure to clinorotation and hyper-g75 and enzyme activities in artemia cysts after space flight76. IAA5 and PIN7 responded to HGMF both under flight (PIN7) and clinorotation (IAA5) and underscores their relevance for auxin modification of growth. Although the number of roots that curved in response to HGMF was low (Table 2), it is possible that differential growth (curvature in response to hydrotropism?) affected the transcription of these genes. GLK did not respond significantly; changes were not limited to any particular condition and therefore cannot be associated with specific experimental conditions but likely represent natural variations. UBq1, Tub1, ADH1, HXK, PFK, SUS, PIN1 showed no significant changes in transcription and therefore could all serve as references. The lack of response of ADH1 is surprising as this gene was previously identified as space stress indicator77,78. TAGL is the only gene that shows reduced transcription in the presence of magnetic fields but only in clinorotated samples. While there is precedence that the lipid metabolism is affected by hypergravity79 and clinorotation80, this is the first observation that magnetic fields might contribute to such changes.
In summary, our data support plant proprioception of weightlessness and metabolic control of amyloplast size. The adjustment of size and mass of amyloplasts indicates that plants perceive gravity and ponderomotive forces, which provide not only enhance gravisensitivity but also explain some metabolic responses to space conditions. The advantage of a sealed environment suggests that this effect is not an artefact but related to gravitational and mechanosensory response. The results also demonstrate that static magnetic fields do not influence gene transcription. Future work needs to investigate the role of other starch-related genes to understand the entire dynamic of metabolic plasticity that relates to weightlessness.