Peroxisomal catalase deficiency modulates yeast lifespan depending on growth conditions

Peroxisomal metabolism triggers lifespan changes in catalase deficient H. polymorpha

To analyze the role of peroxisomal catalase in the CLS of H. polymorpha, CLS measurements were performed using WT and cat strain grown on different carbon and nitrogen sources. Survival measurements started (T = 0 h) when the cultures had entered the stationary phase (Fig. S1ABCD). Very similar CLS curves and mean and maximal lifespans were observed for WT and cat culture grown on glucose/ammonium sulfate (Glc/AS) or glycerol/ammonium sulfate (Gly/AS) (Fig. 1AB, Table 1).

Figure 1. Chronological lifespan of H. polymorpha WT and cat cells grown on different carbon and nitrogen sources. Cells were grown on media containing Glc/AS (A), Gly/AS (B), Glc/MA (C) or Gly/MeOH/AS (D). Data represent mean ± SD, n = 3.

When we used glucose/methylamine media (Glc/MA), catcultures showed a significant reduction in both median and maximum lifespan compared to the WT control (Fig. 1C, Table 1). This reduction was not related to growth differences, as the doubling times and growth yields of both cultures were identical (Fig. S1B).

Upon growth on glycerol/methanol/ammonium sulfate (Gly/MeOH/AS) the median lifespans were similar, but the maximum lifespan of catcultures was significantly enhanced relative to the WT control (Fig. 1D, Table 1). This effect was not related to differences in methanol consumption in both cultures (Fig. S1E). The final optical density of the cat culture on Gly/MeOH/AS was however lower relative to that of WT culture, as reported previously [26] (Fig. S1D).

Catalase deletion affects intracellular ROS levels

To check the impact of CAT deletion on ROS levels, we determined the levels of these compounds using the fluorescent dye dihydrorodhamine 123 (DHR) and FACS (Fig. 2). As shown in Figure 2A cells in the exponential growth phase on Glc/AS, Gly/AS and Glc/MA showed very similar, low ROS levels. Upon growth of WT cells on Gly/MeOH/AS ROS levels were slightly increased. However, in cat cells grown on Gly/MeOH/AS ROS levels were strongly enhanced relatively to that in cat cells grown on the other substrates as well as compared to the WT control grown on Gly/MeOH/AS (>13 ×).

During chronological aging ROS levels increased in all cultures with time. No differences in ROS levels were observed during chronological aging of Glc/AS or Gly/AS grown WT and cat cultures (Fig. 2BC). However, in cat cultures grown on Glc/MA and Gly/MeOH/AS ROS levels increased to higher levels compared to the WT controls (Fig 2DE).

In the Gly/MeOH/AS culture the ROS levels were lower at day 1 (40 h after inoculation; Fig. 2E) relative to the levels observed in the exponential cultures (16 h after inoculation; Fig. 2A) most likely related to the fact that in the stationary phasemethanol oxidation had ceased (Fig. S1E).

Activation of stress adaptation pathways

The high ROS levels observed during the exponential growth phase of Gly/MeOH/AS cat cultures may have induced stress response genes, which are often linked to lifespan extension [27-29]. To test this we analyzed the resistance of WT and cat cells to externally added H₂O₂ or acrolein. Acrolein is a toxic byproduct of lipid peroxidation, which occurs in living cells under conditions of oxidative stress.

Figure 2. Changes in ROS levels in WT andcat cells during exponential growth and chronological aging. DHR staining and FACS analysis was performed with exponentially growing cells (4 h after inoculation for Glc/AS, Glc/MA, Gly/AS and 16 h for Gly/MeOH/AS) (A) and chronologically aging cells grown on media containing Glc/AS (B), Gly/AS (C), Glc/MA (D) or Gly/MeOH/AS (E). Data represents mean ± SD, n = 3. * - P<0.05, ** - P<0.01 in student T test.

Independent of the cultivation conditions used, cat cells displayed decreased resistance to externally added H₂O₂ in comparison to the WT (Fig. 3A), which most likely can be fully explained by the absence of catalase activity in these cells. However, cat cells grown on Gly/MeOH/AS were more resistant to H₂O₂ than Glc/AS, Glc/MA or Gly/AS grown cat cells, suggesting that a catalase independent H₂O₂ defense pathway is induced in these cells. Both WT and cat cultures showed lowest acrolein resistance upon growth on Glc/AS, whereas the resistance slightly increased for both strains upon growth on Glc/MA or Gly/AS, and further increased upon growth on Gly/MeOH/AS. Interestingly, cat cells grown on Gly/MeOH/AS showed higher acrolein resistance than WT cells (Fig. 3B).

Figure 3. H₂O₂ and acrolein resistance of stationary phase WT and cat cells grown on different media. Cells were grown to the stationary phase (16 h on Glc/AS, Glc/MA or Gly/AS and for 40 h on Gly/MeOH/AS). Equal amounts of cells were treated with increasing concentrations of H₂O₂(A) or acrolein (B). Experiments were repeated at least 3 times. Representative plates are shown.

Identification of H. polymorpha Yap1

In S. cerevisiae adaptation to H₂O₂ and thiol reacting compounds, such as acrolein, is mediated by the basic leucine-zipper transcription factor Yap1 [30, 31]. We therefore analyzed the role of H. polymorpha Yap1. The H. polymorpha YAP1 gene was identified using BLAST searches in the H. polymorpha genome database using Yap1 sequences from P. pastoris[32], S. pombe[33] and S. cerevisiae[34]. The putative H. polymorpha Yap1 homologue showed 35% sequence identity to P. pastoris Yap1, 25% to S. pombe Yap1 and 24% to S. cerevisiae Yap1. YAP1 (accession number EFW96135) encodes a protein of 420 amino acids containing a predicted nuclear localization signal (NLS) [35] at the N-terminus (amino acids 30-50) and a leucine rich nuclear export signal (NES) [36] at the C terminus (amino acids 386-392).

Moreover, it contains 6 cysteines, two of which are present within the NES. Sequence alignments indicated that those features are conserved (Fig. S2).

Yap1 is involved in H₂O₂ and acrolein resistance of H. polymorpha cat cells

We first analyzed the role of Yap1 in stress resistance using yap1 and cat yap1 deletion strains. These strains showed similar growth profiles as cat and WT cultures (compare Fig. S1) in media containing Glc/AS, Glc/MA or Gly/AS (Fig. S3ABC). In Gly/MeOH/AS medium, cat yap1 cells showed a slightly enhanced doubling time and a reduced yield relative to cat cultures (Fig. S3D; compare Fig. S1D). Analysis of the resistance to H₂O₂ and acrolein, revealed that this was reduced in cat yap1 cells relative to the cat controls (Fig. 4AB).

Figure 4. Role of Yap1 in H₂O₂ and acrolein resistance and CLS extension. Expotentially Glc/AS growing cat and cat yap1 cells were challenged with increasing concentrations of H₂O₂(A) or acrolein (B). Experiments were repeated twice and a representative experiment is shown. CLS curves of yap1 and cat yap1 strains grown on Glc/AS (C), Gly/AS (D), Glc/MA (E) or Gly/MeOH/MA (F). Number of colonies obtained after 16 h was set to 100% viability except for cat yap1 cells for which the initial viability was measured immediately after shifting to Gly/MeOH medium (inset). Data represents mean viability ± SD, n = 3. (G) Fluorescence microscopy of mGFP-Yap1 complemented yap1 and cat yap1 strains before (0 h) and two hours after the shift to Gly/MeOH/AS media. Brightfield images were false-colored into blue to mark cell borders. Bar - 3μm. (H) Percentage of cells showing mGFP-Yap1 concentrated in a spot after shifting mGFP-Yap1 producing yap1 and cat yap1 strains to Gly/MeOH/AS.

Yap1 is required for lifespan extension of Gly/MeOH/AS grown cat cells

Next, we tested whether Yap1 plays a role in CLS determination. No strong differences in CLS between yap1 and cat yap1 cultures were observed upon growth on Glc/AS or Gly/AS (Fig. 4CD). Moreover, the curves were similar to those observed for WT and cat (Fig. 1). On Glc/MA, the yap1 curve resembled that of WT, whereas the cat1 yap1 survival curve was similar to that of cat cells (Fig. 4E, compare Fig. 1C), indicating that Yap1 is not an important player in determining the CLS at these conditions. The CLS of yap1 cultures on Gly/MeOH/AS was comparable to that of the WT control (Fig. 4F, compare Fig. 1D). In contrast a very rapid decrease in survival was observed for Gly/MeOH/AS grown cat yap1 cells resulting in less than 10% survival within 12 h after shifting the cells to this medium (Fig. 4F, inset). These data indicate that Yap1 is of major importance for survival of Gly/MeOH/AS grown cat cultures.

To seek further evidence for the function of Yap1, we performed localization experiments using a gene encoding an N-terminal fusion of mGFP with Yap1 under control of the YAP1 promoter. Growth and CLS experiments revealed that the fusion protein fully complemented the yap1 deletion strain (data not shown). Upon growth on Glc/AS, Glc/MA or Gly/AS, mGFP-Yap1 was predominantly localized to the cytosol in WT and cat cells (data not shown). However, after the shift of Glc/AS-grown cells to Gly/MeOH/AS, mGFP-Yap1 rapidly migrated (within two hours) to the nucleus of cat cells, but not of WT cells (Fig. 4GH). These data reveal rapid Yap1 activation upon exposure of cat cells to methanol.

Cytochrome c peroxidase and superoxide dismutase, but not glutathione reductase, are induced in Gly/MeOH/AS grown cat cells

Induction of antioxidant enzymes has been implicated in yeast lifespan extension. We determined the activities of cytochrome c peroxidase (CCP), superoxide dismutase (SOD) and glutathione reductase (GLR) in crude extracts of WT and cat cells grown for 16 h on Glc/MA or Gly/MeOH/AS. As shown in Figure 5A, CCP is induced 3-fold in Gly/MeOH/AS-grown cat cells relative to the WT control. During growth on Glc/MA CCP activities were similar in both strains (Fig. 5A).

To check whether induction of CCP depends on Yap1, we also measured CCP activities in Gly/MeOH/AS grownyap1 and cat yap1 cells, using WT and cat cells as controls. Deletion of YAP1 alone did not affect activity of this enzyme relative to WT. CCP activity was enhanced in cat cells but not in cat yap1 cells relative to the WT control (Fig. 5B).

SOD isozyme profiling of stationary WT cells grown on Glc/AS medium revealed the presence of 3 bands(Fig. 5C). The bands of lowest (designated SOD1) and highest (designated SOD3) relative mobility were inactivated by H₂O₂ and KCN and therefore represent Cu/Zn containing SOD enzymes. The middle band (designated SOD2) is most likely a Mn containing SOD, as it appeared to be resistant to both compounds.

Comparison of isozyme profiles of WT and cat cells grown on Glc/MA or Gly/MeOH/AS, revealed that cat cells grown on Gly/MeOH/AS showed the highest SOD1 activity (Fig. 5D), which was inhibited by H₂O₂ and KCN (Fig. 5E).

Because cat yap1 cells very rapidly die on Gly/MeOH/AS media (Fig. 4F), SOD activities were determined in these cultures obtained 4 hours after the shift from Glc/AS to Gly/MeOH/AS. At this time point the activity of SOD1 had not yet increased in cat cells. In cat yap1 cells however SOD levels were increased (Fig. 5F), suggesting that SOD1 expression is not controlled by Yap1.

Enzyme activities of GLR were similar in WT and cat cells grown on Glc/MA or Gly/MeOH/AS (Fig 5E).

Together our data indicate that in cat cells grown on Gly/MeOH/AS, but not on Glc/MA, genes are expressed that enhance stress resistance and extend the CLS. One of these genes encodes CCP and is induced in a Yap1 dependent way, another one is SOD1, which is not under control of Yap1.

CCP plays a role in lifespan extension of cat cells grown on Gly/MeOH/AS

The increase in lifespan of cat cells on Gly/MeOH/AS depends on Yap1 and is paralleled by an increase in CCP activity, which also depends on Yap1. To test whether indeed enhanced CCP levels are essential for lifespan extension, CLS experiments were performed with a ccp deletion strain and a cat ccp double deletion strain. As shown in Figure 5H, CCP is essential for survival ofcat cells in Gly/MeOH/AS medium. Deletion of CCP alone did not have a significant effect on CLS (Fig. 5H). These data suggest that in cat cells in Gly/MeOH/AS medium Yap1 mediated induction of CCP, is essential for lifespan extension.

Figure 5. Cytochrome c peroxidase, superoxide dismutase and glutathione reductase activities in H. polymorpha WT and cat cells grown on different media. (A) Measurements of CCP activities in crude extracts of WT and cat grown on Glc/MA and Gly/MeOH/AS for 16 h. (B) Measurements of CCP activities in crude extracts of WT, cat, yap1 and cat yap1 cells grown on Gly/MeOH/AS for 4 h. (C) Superoxide dismutase isozyme profiling in WT cells. SOD activity was detected in crude extracts prepared from cells grown for 16 h on Glc/AS without pre-treatment (NT) or upon 30 min pre-incubation with 5 mM H₂O₂ or 5 mM KCN. (D) SOD activity in WT and cat cells grown for 16 h on Glc/MA or Gly/MeOH/AS. (E) SOD activity of in cat cells grown for 16 h in Gly/MeOH/AS detected without pre-treatment (NT) or after 30 min pre-incubation with 5 mM H₂O₂ or 5 mM KCN. (F) SOD activities in WT, cat, yap1, cat yap1 cells grown for 4 h on Gly/MeOH/AS. (G) Glutathione reductase activities in WT and cat cells grown in Glc/MA or Gly/MeOH/AS for 16 h. (H) CLS of ccp and cat ccp cells in Gly/MeOH/AS medium. Viability experiments were started 12 h after inoculation of the media. Experiments were repeated at least twice. Data represent mean activities ± SD, n = 3, ** - p<0.01. For native gels, representative gels are shown.

AMO activity is partially reduced in Glc/MA grown cat cells

We recently showed that the presence of MA instead of AS as sole nitrogen source in H. polymorpha cultures extends the CLS, because MA serves as extra energy source in stationary phase cultures [37]. Since the lifespan of cat cells in Glc/MA is reduced relative to the WT control, but similar to that of Glc/AS cultures (Fig. 1AC) we investigated whether this is related to the absence of amine oxidase (AMO) activities, which would prevent the generation of energy from MA during chronological aging. As shown in Figure 6, AMO activity was still present albeit reduced by 53% in cat cells in comparison to WT in stationary cultures (Fig. 6A). However, the residual AMO activity most likely still allows to metabolize MA during the stationary phase. Western blot analysis revealed that AMO protein levels were only slightly reduced, indicating that part of the AMO enzyme is inactivated incat cells (Fig. 6BC).

Figure 6. AMO activities are reduced in Glc/MA cat cells relative to the WT control. (A) Measurements of AMO activities in crude extracts of WT and cat cells grown on Glc/MA. Activity is expressed as mU/mg protein. (B) Western blot analysis of AMO protein levels in cells grown for 16 h on Glc/MA. Cytosolic pyruvate decarboxylase Pyc1 was used as a loading control. (C) Quantification of AMO protein relative to Pyc1 of the blot shown in B. Data represent mean ± SD, n = 3. (D) CLS of WT and cat cells grown on Gly/MeOH/MA. Time point 0 h indicates time when cells were shifted to final medium. The number of colonies obtained after 16 h of growth was set to 100% viability. Experiments were repeated twice. Data represents mean ± SD, n = 3.

The activation of stress responsive genes does not alleviate the negative effect of MA on the CLS of cat cells

Our data suggested that upon growth on Gly/MeOH/AS, but not on Glc/MA, resistance genes are induced in cat cells. To test whether the induction of these genes can overcome the negative effect of MA on cell survival, we compared the CLS of WT and cat cells grown on Gly/MeOH supplemented with MA as sole nitrogen source. In agreement with our previous data addition of MA extended the lifespan of WT cells, however cat cultures showed a strongly reduced mean and maximum lifespan relative to the WT control (Fig. 6D)

Strains and growth conditions

The H. polymorpha strains used in this study are listed in Table 3. Cells were grown on mineral medium (MM) [44] supplemented with different carbon sources: 0.25% glucose (Glc), 0.05% glycerol (Gly) or a mixture of 0.05% glycerol and 0.5% methanol (Gly/MeOH), and nitrogen sources: 0.25% ammonium sulfate (AS) or 0.25% methylamine(MA). 6 mM K₂SO₄ was added when methylamine was used as nitrogen source. When required, leucine was added to a final concentration of 60 μg/ml. Selection of yeast transformants was performed on YND plates (0.17% Yeast Nitrogen Base w/o AA, w/o N, 0.25% NH₄SO₄, 1% glucose and 2% agar) or on YPD (1% yeast extract, 1% glucose, 1% peptone, 2% agar) supplemented with 300 μg/ml hygromycin B (Sigma) or 100 μg/ml nourseothricin. For viability determination cells were plated on YPD agar plates. For cloning purposes, E. coli DH5α or GM48 were used. Bacteria were grown at 37°C in LB media supplemented with 100 μg/ml ampicillin or 50 μg/ml kanamycin when required.

Cloning and construction of yeast strains

The plasmids and primers used in this study are listed in Tables 4 and 5. All cloning was performed using Gateway technology (Invitrogen). Standard recombinant DNA techniques and transformation of H. polymorpha was performed as described previously [45]. All deletions and integrations were confirmed by PCR and southern blotting.

Construction of a H. polymorpha cat deletion strain

A catalase deletion strain (cat) was constructed by replacing the genomic region of CAT (P30263) comprising nucleotides +1 to +1256 by the auxotrophic marker for uracil (URA3). To this end the first region -399 to 0 of the CAT gene was amplified using primers 41_5CAT_F and 41_5CAT_R with attB sites and recombined into pDONR41 yielding pENTR_41_5'CAT. At the same time region +1256 to +1665 was amplified using primers 23_3CAT_F and 23_3CAT_R and recombined with pDONR23 yielding pENTR_23_3'CAT.

A deletion cassette containing 5’ and 3’ fragments of the CAT gene and the URA3 marker was assembled in pDEST43 with Gateway LR reaction using plasmids pENTR_41_5'CAT, pENTR_221_URA3 and pENTR_ 23_3'CAT. The resulting plasmid pDEL_CAT_URA3 was used as a template to amplify a deletion cassette of 2595 bp in PCR reaction using primers Catdelcas_F and Catdelcas_R. The purified PCR product was transformed into H. polymorpha NCYC 495 ura3 leu1.1 and colonies were selected on YND with leucine.

Construction of H. polymorpha yap1 and cat yap1 strains

YAP1 (EFW96135) was deleted by replacing nucleotides -2 to +1262 by a hygromycin B resistance cassette (HPH). To this end a fragment containing region -280 to -3 from the start codon was first amplified from H. polymorpha genomic DNA using primers 41_YAP1_F and 41_Yap1_R and recombined in Gateway BP reaction into pDONR_41 yielding plasmid pENTR_41_5'YAP1. Similarly region +1263 to +1639 was amplified using primers 23_YAP1_F and 23_YAP1_R and recombined into pDONR_23 yielding pENTR_23_3'YAP1. Plasmids pENTR_41_5'YAP1, pENTR_221_HPH and pENTR_23_3'YAP1 were recombined in Gateway LR reaction with pDEST_43_ NAT yielding plasmid pDEL_YAP1_HPH. A deletion cassette of 2417bp was amplified from this plasmid using primers Yapdelcas_F and Yapdelcas_R and used for transformation of H. polymorpha WT and cat cells. Transformants were selected on YPD with hygromycin.

Complementation of yap1 with mGFP-YAP1

The YAP1 deletion strain was complemented by insertion of an expression cassette containing the YAP1 promoter, followed by a gene encoding the N terminally mGFP tagged Yap1 and the AMO terminator into the YAP1 promoter region. To this end first the YAP1 promoter (region -400 to 0) was amplified from genomic DNA using primers 41_Prom_Yap1F and 2PGFP_O_R. The mGFP ORF without a stop codon was amplified from the pHIPZ_mGFP fusinator plasmid using primers 2PGFP_O_F and 41_mGFP_R. PCR fragments were combined and used as a template in second overlay PCR using primers 41_Prom_Yap1F and 41_mGFP_R. The obtained DNA fragment was used in a Gateway BP reaction with pDONR_41 yielding pENTR_41_ P_YAP1_mGFP. Next YAP1 ORF was amplified from genomic DNA using primers 221_YAP1_F and 221_YAP1_R and recombined in Gateway BP reaction with pDONR_221 resulting in plasmid pENTR_221_YAP1. Plasmids pENTR_41_P_YAP1_mGFP, pENTR_221_YAP1, pENTR_23_T_AMO were recombined in a Gateway LR reaction with pDEST_43_NAT yielding plasmid pEXP_P_YAP1_mGFP-YAP1_T_AMO. This plasmid was transformed to E.coli GM48 to obtain unmethylated DNA. After linearization with XbaI plasmid DNA was transformed into yap1 and cat yap1 cells. Colonies were selected on YPD supplemented with nourseothricin.

Construction of H. polymorpha ccp and cat ccp strains

CCP (EFW94326) was deleted by replacing nucleotides -7 to +376 by a HPH cassette. Region -280 to -3 from the start codon was first amplified from H. polymorphagenomic DNA using primers 41_CCP_F and 41_CCP_R and recombined in Gateway BP reaction into pDONR_41 yielding plasmid pENTR_41_5'CCP. Similarly, region +376 to +795 was amplified using primers 23_CCP_F and 23_CCP_R and recombined into pDONR_23 yielding pENTR_23_3'CCP. Plasmids pENTR_41_5'CCP, pENTR_221_HPH and pENTR_23_ 3'CCP were recombined in a Gateway LR reaction with pDEST_43_NAT yielding plasmid pDEL_CCP_HPH. A deletion cassette of 2633 bp was amplified from this plasmid using primers Ccpdelcas_F and Ccpdelcas_R and used for transformation of H. polymorpha WT and cat cells. Transformants were selected on YPD with hygromycin.

Chronological lifespan measurements

Cells were extensively precultivated on media containing 0.25% glucose and 0.25% ammonium sulfate. Mid-exponential cultures (OD_600nm = 1.8) cultures were diluted to an OD_{600 nm} of 0.1 in the final medium. Survival measurements were started after the cultures reached the stationary phase. This was 16 h for cultures on Glc/AS,Glc/MA and Gly/AS and 40 h for cultures grown on Gly/MeOH/AS unless stated otherwise. The number of cells per ml of culture was determined using CASY^® Model TT (Roche Applied Science). 500 cells were plated on YPD agar plates in triplicate. Plates were incubated at 37°C for 36 to 48 hours and photographed. Colony numbers were counted using an ImageJ plugin. The number of colony forming units at the first time point (invariably approximately 500) was set to 100%.

Stress resistance experiments

WT and cat cells grown on for 16 h on 0.25% glucose / ammonium sulfate, 0.25% glucose / 0.25% methylamine, 0.05% glycerol / ammonium sulfate or 40h for 0.05% glycerol / 0.5% methanol / ammonium sulfate were collected by centrifugation and resuspended in 50mM potassium phosphate buffer pH=6.0 to an OD_{600 nm} of 0.3. 1 ml of the suspension was challenged with increasing concentrations of H₂O₂ or acrolein for 1 hour at 37°C. Cells were washed once and resuspended in 1 ml 50 mM potassium phosphate buffer pH=6.0. 5μl of the suspension was spotted on YPD agar plates and plates were incubated 24 h at 37°C.

Biochemical methods

Preparation of cell extracts [46], determination of methanol concentrations [47], enzyme assays for AMO [48], superoxide dismutase [49,50] and cytochrome c peroxidase [51] were performed as described earlier. Glutathione reductase activity was determined using the glutathione reductase assay kit (Sigma-Aldrich). Protein samples for SDS-PAGE gels were prepared as described before [52]and separated on 10% polyacrylamide gels. Proteins were transferred to nitrocellulose membranes using the semi-dry blotting method and probed with specific polyclonal rabbit anti-AMO or anti-pyruvate carboxylase (Pyc1; loading control) antibodies. Blots were quantified using ImageJ.

ROS measurements

ROS accumulation was measured using dihydrorhodamine 123 (DHR, Invitrogen). 10⁷ cells were harvested and stained in 50 mM potassium phosphate buffer pH=7.0 containing 20 μg/ml DHR for 30 minutes at room temperature in the dark. Cells were washed once and resuspended in the same buffer. Cells treated in same way without dye were used as controls for fluorescence background. Fluorescence signal of individual cells was captured in a FACS Aria II Cell sorter (BD Biosciences) for 10.000 events at the speed of 500-1000 events per second using a 488nm laser, 505nm long pass mirror and 525/50nm band-pass filter. FACSDiva software version 6.1.2 was used for data acquisition and analysis. The presented data represent differences in mean fluorescence between stained cells and the background.

Fluorescence microscopy

Fluorescence microscopy images were captured using a Zeiss Axioskop 50 with a 100x 1.30 NA Plan Neofluar objective using MetaVue software and a digital camera (model 1300Y; Princeton Instruments). GFP signal was visualized with a 470/40 nm bandpass excitation filter, a 495 nm dichromatic mirror, and a 525/50-nm bandpass emission filter. ImageJ and Adobe Photoshop CS2 were used for image analysis and figure preparation. In overlay figures bright field images were false colored in blue to mark cell edges.