SEARCH

SEARCH BY CITATION

Keywords:

  • Yarrowia lipolytica;
  • alkaline extracellular protease;
  • protein transport;
  • secretory pathway;
  • biological stress

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Upon transfer to alkaline extracellular protease (AEP) induction medium, strain 773-2 (50 integrated copies of XPR2), derived from highly inbred strain E129, grew for at least 10 h before AEP production began, and then growth rate decreased before increasing again; by then, cells had lost copies of XPR2 (Le Dall et al., 1994). Slowing of growth following AEP induction suggested that increased secretory pathway cargo load was affecting cell growth and that such a system had potential for secretion stress studies. Development of W29-derived XPR2 multi-copy strains and improved AEP induction conditions realized this potential. AEP production was sixfold higher than for 773-2. Rapid AEP induction and slowing of growth by 3 h minimized loss of XPR2 gene copies. Two strains, examined in more detail, differed in initial AEP productivity, extent of slowing of growth during AEP induction, and subsequent recovery of growth rate and AEP productivity demonstrating that the system provides a range of secretion stresses and ensuing adaptations. W29-derived strains should be more ‘wild type’ than 773-2 for secretory pathway components and their regulation. They should provide an excellent system for kinetic analysis of gene expression responses to acute increases in secretory pathway cargo load.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Induction of XPR2 coding for alkaline extracellular protease (AEP) in Y. lipolytica increases secretory load. Induction in strains with multiple copies of XPR2 increases secretory load rapidly enough that cells cannot completely adapt and growth rate slows (Le Dall et al., 1994), suggesting that such strains might be developed into an excellent system for kinetic analysis of the secretion stress response.

Gene expression analysis might reveal differences between Y. lipolytica and Saccharomyces cerevisiae in regulatory responses to increases in secretory pathway cargo load. By several criteria, Y. lipolytica is a far better secretor (Dalboge, 1997; Madzak et al., 2004). Saccharomyces cerevisiae thrives in an environment with high simple sugar content, and it secretes only low levels of hydrolases (Martini, 1993). In contrast, Y. lipolytica is isolated from diverse environments where it can depend on hydrolases secreted into the extracellular medium such as proteases, lipases, phosphatases, and RNases to supply nutritional needs (Barth & Gaillardin, 1996). Higher biases in codon usage values for secretory pathway genes for Y. lipolytica than for S. cerevisiae (see 'Results') suggest that these components could be expressed at higher levels in Y.  lipolytica. Possibly, rather than constitutively synthesizing substantial excess secretory capacity, Y. lipolytica might up-regulate the levels of secretory pathway components when it must secrete large amounts of enzymes to compete and survive. Saccharomyces cerevisiae does not face this challenge and might not regulate these components as extensively.

In S. cerevisiae, when the cargo load exceeds the folding capacity of the endoplasmic reticulum (ER), unfolded protein response (UPR) is induced (Patil & Walter, 2001). UPR results in ER membrane expansion, and induction of some genes encoding components involved in protein folding in the ER, in ER-associated degradation (ERAD), and in protein translocation and transport into and through the secretory pathway (Schuck et al., 2009). Only a few of the translocon components and none of the targeting components are induced in S. cerevisiae during UPR (Travers et al., 2000). UPR is more complex in some other organisms. UPR in mammalian cells shares the IRE1 signal transduction pathway and the Hac1p/XBP-1 transcriptional activators with S. cerevisiae, and it includes two additional pathways with ATF6 and PERK as the sensors (Ron & Walter, 2007).

In addition to UPR, other responses to secretory load have been identified in other organisms. Increased protein load entering the cis-Golgi from the ER in mammalian cells has been shown to up-regulate secretion without increasing levels of secretory pathway components (Pulvirenti et al., 2008). Levels of secretory pathway components are up-regulated during development of cells destined to be professional secretors such as B lymphocytes that differentiate into Ig-secreting plasma cells. A proteomic approach revealed that the secretory machinery expands before production of IgM increases (van Anken et al., 2003). The initial expansion does not appear to involve classical XBP-1-mediated UPR. CreB/Cre3-like transcription factors are major and direct regulators of secretory capacity (Fox et al., 2010). Drosophila CrebA directly binds the enhancers of secretory pathway genes and is both necessary and sufficient to activate expression of every secretory pathway component gene examined thus far. Human CrebA orthologs have the ability to regulate the secretory pathway in nonsecretory cell types.A multi-gene approach combining Sly1, Munc18c, and constitutively expressed XBP-1 in transgenic mammalian cells secreting recombinant proteins seems to improve a cell's ability to secrete increased cargo loads and to overcome limitations on high production caused by saturation of the secretory pathway (Peng & Fussenegger, 2009).

The improved Y. lipolytica XPR2 multi-copy strain system described here should provide a direct approach to studying responses to secretory pathway overload by overexpression of AEP, a homologous secreted protein, and determination of the transcriptional or proteomic responses. Examples of such studies with yeast, usually involving overexpression of a heterologous protein, include S. cerevisiae (Casagrande et al., 2000), Kluyveromyces lactis (van Ooyen et al., 2006), and Pichia pastoris (Gasser et al., 2006; 2007). Examples with filamentous fungi include Trichoderma reesei (Arvas et al., 2006) and Aspergillus niger (Guillemette et al., 2007; Jorgensen et al., 2009).

This study describes construction and characterization of improved Y. lipolytica XPR2 multi-copy strains and improvements in AEP induction conditions that make rapid increases in secretory cargo load possible. These improvements overcome many of the difficulties that the previously described Y. lipolytica E129-derived XPR2 multi-copy strain system (Le Dall et al., 1994) would present for doing gene expression studies. These include loss of XPR2 gene copies, complications in determining when to take samples to capture the increases in secretory cargo load, and inability to grow cells in defined preculture medium with NH4+ as N source where AEP is repressed. W29-derived strains secrete much higher levels of heterologous proteins (Dalboge, 1997) and AEP (this study) than E129-derived strains suggesting that extensive inbreeding may have adversely affected secretory abilities of E129 strains and that W29-derived strains would be much closer to ‘wild type’ for secretory pathway components and secretory pathway regulation. Therefore, W29-derived strains would be a much better strain choice for gene regulation studies of responses to acute increases in secretory pathway cargo load.

Growth and AEP production/productivity of new XPR2 multi-copy strains were screened and two were examined quantitatively in more detail. The strains differed in the extent of secretory pathway cargo overload, growth responses, and ensuing recovery of growth rate and AEP productivity. This system can provide a range of intensities of secretory stress and responses, and it should be excellent for kinetic analysis of changes in gene expression during adaptation to acute increases in secretory pathway cargo load.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Strains and media

Escherichia coli JM105 was used for DNA manipulations (Sambrook et al., 1989). Yarrowia lipolytica strains were derived from CX161-1B, W29, and E129 parental strains (Fig. 1). CX161-1B is from an inbreeding program in R. Mortimer's laboratory started with YB423-3 x YB423-12 (Barth & Gaillardin, 1996). CX161-1B Δura3 (MATA ade1 ura3E), previously designated SMY1, has an EcoRV deletion of URA3 (Matoba et al., 1997). W29 is a wild-type strain (Gaillardin et al., 1973). W29-Δura3 has ura3-302, an internal and almost complete deletion of URA3 (Barth & Gaillardin, 1996). E129 is from an inbreeding program in H. Heslot's laboratory started with W29 x YB423-12 (Barth & Gaillardin, 1996). 20-12 (MATA ura3-302 leu2-270 xpr2-322 [XPR2]) is E129 with a single copy of XPR2 integrated, and 773-2 is E129 with almost 50 copies of XPR2 integrated (Le Dall et al., 1994).

image

Figure 1. Relationships of strains used in this study.

Download figure to PowerPoint

CX161-1B Δura3 and W29-Δura3 strains with multiple copies of XPR2 were constructed by transformation with plasmid pINA773. It contains a fragment of the Y. lipolytica rDNA, a defective version of the URA3 gene (most of the promoter deleted), a pBR322 fragment for selection and replication in E. coli, and the XPR2 gene. Transformation with pINA773 linearized in the rDNA fragment resulted in transformants with 25–60 copies of XPR2 integrated into genomic rDNA sites (Le Dall et al., 1994). Transformations were carried out with 5 mg of KspI-cut pINA773 (Le Dall et al., 1994). Plasmid DNA was isolated using Qiagen Hi Speed Midi Preps (Valencia, CA). Transformants were selected after 5 days and streaked on YLT-ura plates.

Four 30-mL solutions of proteose peptone (2% w/v) were dialyzed (6000–8000 MW cut off) three times against 4 L of water at 4 °C to produce dialyzed proteose peptone; samples were stored at −20 °C. YLT-ura is a synthetic complete medium lacking uracil (Matoba et al., 1997). SC-ura.2 and SC-ura.3 are defined media, in which AEP production is repressed, used to grow cells for AEP induction experiments. SC-ura.2 contains glycerol (10 g L−1), yeast nitrogen base (YNB) without amino acids and ammonium sulfate (1.7 g L−1), ammonium sulfate (1.0 g L−1), and SC-ura dropout mixture (600 mg L−1), prepared as described by Davidow et al. (1985) with 10 instead of 20 g of threonine. SC-ura.3 contains glycerol (10 g L−1), YNB without amino acids and ammonium sulfate (0.85 g L−1), ammonium sulfate (1.0 g L−1), SC-ura dropout mixture (600 mg L−1), and 50 mM phosphate buffer (pH 6.8). To produce DPP, three solutions – DDPI, DPPII, and 0.5 M phosphate buffer (pH 6.8) – were autoclaved separately and 10, 30, and 10 mL of each combined per flask. DPPI contains 54 mL of dialyzed proteose peptone (originally 2% w/v) and 6 mL of 2% proteose peptone in 120-mL final volume. Compared with 100% dialyzed proteose peptone, this 90%:10% mixture reduced the growth lag after transfer and still allowed rapid AEP induction. DPPII contains 13.4 g of 50% glycerol, 1.66 g YNB without amino acids and ammonium sulfate, and 30 mg adenine in 600 mL final volume.

Codon bias

Codon adaptation index (CAI) (Sharp & Li, 1987) for Y. lipolytica genes was calculated using a reference table of relative synonymous codon usage based on the seven genes with the most highly biased usage of the first 77 Y. lipolytica genes sequenced – POT1, TEF1, RPS7, PGK1, KAR2, ICL1, and ACO2. CAIs of S. cerevisiae genes were from the Saccharomyces Genome Database (www.yeastgenome.org).

Temperature effects on AEP production

CX161-1B was grown in GPP medium (Ogrydziak & Scharf, 1982) containing L−1 100 mg adenine, 50 mg gentamycin sulfate, and 0.2 mL polypropylene glycol (MW 2000, Polyscience) in a 14 L automated New Brunswick Microferm Bench Top Fermentor. Agitation speed was 800 r.p.m. and dissolved oxygen > 70%; pH was controlled at 6.8, and temperatures at 19.9, 23.0, 25.1, 27.7, 30.0, or 32.8 °C. AEP activity was measured by p-tosyl-l-arginine-methyl-ester hydrolysis and cell mass determined from cell dry weight.

Screening for AEP production on plates

Transformants were prescreened by streaking twice on YLT-ura plates. Patches grown on YLT-ura plates were used to inoculate SC-ura3 tubes. For each transformant, one SKM (skim milk) + ade plate (Enderlin & Ogrydziak, 1994) was spotted with three separate 5 μL drops (103 transformants per drop). Plates were incubated at 23 °C and diameters of colonies and zones of clearing measured to the nearest 0.2 mm every 24 h for 4 days. AEP activity was estimated using the standard curve of diameters of zones of clearing vs. AEP activity found for AEP in an agar diffusion assay in which wells punched in SKM azide plates were filled with known quantities of AEP (Ogrydziak & Scharf, 1982). Colonies were disk shaped, and relative cell numbers were estimated from the square of colony diameter.

Screening for AEP induction in liquid culture

Isolates were grown overnight in tubes with SC-ura2 medium at 28 °C with shaking. Unbaffled flasks (500 mL) with 50 mL of SC-ura3 medium were inoculated with sufficient cells to produce cell densities > 400 Klett units (KU) the next morning. Flasks were incubated at 23 °C with shaking. Cells were collected by centrifugation and resuspended at a calculated 275 KU cell density in 30 mL of DPP medium in a 500-mL baffled flask. Flasks were incubated at 23 °C and 250 r.p.m.. Samples for cell density measurements and AEP assays were taken every 1–2 h for the first 10–12 h and less often afterward. Culture medium pH was measured every 6 h.

Cell density was measured with a Klett meter (green filter). One KU was about 2 × 105 cells mL−1, and 250 KU was about 1 g L−1 dry cell weight. AEP activity was estimated using the agar diffusion assay with SKM azide plates (Ogrydziak & Scharf, 1982).

Cell accumulation on flask walls was a problem in early experiments. Silanizing with Sigma Kote did not help. Machine washing of the flasks, limiting shaking speed to 250 r.p.m., and adding antifoam (50 μL of polypropylene glycol (MW 2000, Polyscience)) at 0 and 12 h largely eliminated the problem.

Calculation of IVCD and cell productivity (qp)

qp was calculated from ΔP/ΔIVCD over the cumulative time interval, tt0 (Adams et al., 2007). ΔP is the change in AEP concentration. IVCD is Integral of Viable Cell Density; it represents the cumulative cell concentration* time during which AEP is produced. IVCD units are 109 cells*h L−1, and IVCD was calculated by (xxt−1)/2*Δt + IVCDt−1. (See Supporting Information, Appendix S1 for details).

About 1 mg of AEP has 9000 units of activity (Ogrydziak & Scharf, 1982). Therefore, one unit is equivalent to 111 ng of AEP. Product concentration of AEP (mg L−1) was calculated by multiplying AEP concentration (U mL−1) by 103 mL L−1 and 111 ng U−1 and 10−6 mg ng−1. One Klett unit is equal to 2 × 105 cells mL−1. For calculation of IVCD, Klett readings were multiplied by 2 × 105 cells mL−1 and 103 mL L−1.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Synonymous codon usage bias suggests higher expression levels of components of the early secretory pathway for Y. lipolytica than for S. cerevisiae

For highly expressed genes, a subset of codons is preferentially used (Sharp & Cowe, 1991). For genes expressed at low levels, usage tends to be more uniform. Synonymous codon usage bias was significantly higher for the first sequenced Y. lipolytica genes involved in protein targeting to the ER, translocation across the ER, and folding in the ER than for the S. cerevisiae homologs. Yarrowia lipolytica and S. cerevisiae CAI values for subunits of the signal recognition particle and its receptor, of translocon components, and of select proteins involved in translocation, including KAR2, are shown in Table S1. In most cases, the Y. lipolytica values are 2- to 3-fold higher suggesting that proteins involved in protein targeting and translocation could be produced at much higher levels in Y. lipolytica.

Developing AEP induction conditions

Alkaline extracellular protease production depends on strain background, growth conditions, and media composition. It was triggered by N limitation, by the presence of peptone but repressed by amino acids (Ogrydziak et al., 1977; Ogrydziak & Scharf, 1982; Le Dall et al., 1994). It was also affected by yeast extract and the specific batch of peptone used (unpublished results). When multi-copy XPR2 strain 773-2, constructed in the E129 background (Fig. 1), was transferred to a glucose proteose peptone medium at pH 6.8, the strains grew fairly rapidly at first. At least 10–11 h after transfer, AEP production was induced and cells then grew more slowly for several hours before resuming somewhat faster growth. During this time, there was selection for cells that had lost copies of the XPR2 gene (Le Dall et al., 1994).

We reasoned that if AEP production could be induced very rapidly before copies of XPR2 could be lost, then AEP induction in XPR2 multi-copy strains would provide an excellent system for examining secretion stress caused by rapid secretory pathway cargo overload. The major change in the induction protocol was the use of dialyzed proteose peptone as the N source. Removal of amino acids and low-molecular-weight peptides that repress AEP production resulted in much faster induction. Another change was that inoculum was grown in selective defined minimal medium (SC – synthetic complete) at pH 6.8 to better maintain the high XPR2 copy number and to avoid the pH change upon transfer to AEP induction medium (see below). Changing to pH 7.2 did not improve induction (data not shown). Glycerol was found to support more rapid growth and slightly higher AEP production and was substituted for glucose in both the SC and induction media.

Previously, AEP induction experiments were carried out at 28 °C (Le Dall et al., 1994). A detailed study on temperature effects on growth rate and AEP production had been carried out for CX161-1B in an instrumented 14L fermentor (data not shown). While the maximum growth rate (μ) was near 28 °C, the maximum AEP volumetric yield (U mL−1) and AEP productivity (U/cell mass) were at 19.9 °C (lowest temperature investigated). Maximum differential productivity (μ*ΔU/Δ cell mass) was around 23 °C, and 23 °C was used for further studies. With these improved AEP induction conditions, growth of 773-2 slowed by 3 h after transfer, much sooner than the more than 10 h originally found (data not shown).

Strain selection

CX161-1B vs. 20-12 and 773-2

CX161-1B, which has one copy of XPR2, grew more rapidly, produced more AEP, and had higher cell yields than 20-12 (E129 derivative with one copy of XPR2). 773-2 (E129 derivative with 40–50 copies of XPR2) produced about 10-fold more AEP than 20-12, but only about twofold more than CX161-1B (data not shown). E129 was extensively inbred for genetic studies. E129-derived strains 20-12 and 773-2 grow poorly with NH4+ as N source. They were kept on YNB plates with glutamate and casamino acids as N sources, and precultures were carried out on YPD at pH 5.0 to prevent AEP induction (M.-T. Le Dall, pers. comm). Preculture in SC medium at pH 6.8 would avoid the complication of gene expression changes owing to the pH shift from 5.0 to 6.8. Because of expected higher AEP production and the ability to grow with NH4+ as N source, XPR2 multi-copy strains were constructed in CX161-1B Δura3.

CX161-1B Δura3 vs. W29 Δura3

W29-derived strains secreted much higher levels of heterologous proteins compared with E129-derived strains (Dalboge, 1997). Also, CX161-1B Δura3 did not grow as rapidly as hoped for in SC medium. Therefore, W29 Δura3 and CX161-1B Δura3 were compared for growth in versions of SC medium and for AEP production in DPP medium.

Synthetic complete (SC) used for the initial transfer experiments was unbuffered (initial pH about 5.7). For W29 Δura3, vigorous initial growth rapidly reduced pH and this lead to decreased cell yield. Including phosphate buffer significantly increased cell yield – OD600 of 11.5 at 17 h. For W29 Δura3, decreasing glycerol from 20 to 10 g L−1 and ammonium sulfate from 5.0 to 1.0 g L−1 had little effect on cell yield at 17 h. For CX161-1B Δura3, these changes increased cell yield from 2.4 to 4.0 (OD600) at 23 h (still much lower than for W29 Δura3). However, at these concentrations, buffering actually reduced cell yield to OD600 of 2.3 at 23 h. In buffered medium, further reducing YNB without amino acids and ammonium sulfate from 1.7 to 0.85 g L−1 increased cell yield of CX161-1B Δura3 to OD600 of 3.6 at 23 h. Based on these results, SC-ura.2 (used for plates and tubes in the first steps of preparing inocula) and SC-ura.3 (used for preculture of cells before transfer to AEP induction medium) were developed as media in which both CX161-1B Δura3 and W29 Δura3 would grow reasonably well.

W29 Δura3 made more AEP than CX161-1B Δura3 (Table S2) and grew faster in the various SC media. Therefore, XPR2 multi-copy strains were also constructed in W29 Δura3.

Screening for AEP production of XPR2 multi-copy transformants on plates

About 200 transformants were streaked twice on YLT-ura plates. Those yielding only small, slow-growing colonies were not examined further. Depending on size and consistency of size, 64 isolates (58 derived from W29 Δura3 and 6 from CX161-1B Δura3) were analyzed further for growth and AEP production on plates.

The diameters of colonies and their zones of clearing on skim milk (SKM plates) were slightly larger for W29 Δura3 compared with CX161-1B Δura3 (Table S2). For W29 Δura3 XPR2 multi-copy transformants, zones of clearing were comparable to those for parental W29 Δura3 but colonies were always smaller, consistent with the XPR2 multi-copy strains producing more AEP per cell (Table S2). Relative AEP production (AEP/cells) of spotted cells was roughly estimated using the diameter of the zone of clearing and square of colony diameter (see 'Materials and methods'). The seven transformants producing the highest levels of AEP were all derived from W29 Δura3 They were designated Y2 to Y69 (Table S2) and examined further in liquid culture.

Screening for AEP induction of XPR2 multi-copy transformants in liquid culture

Y69 was eliminated because of poor growth in SC media. Results for the six other transformants are in Table 1. Compared with W29 Δura3, AEP production for Y3 was 5.9× higher at 12 h and 12.4× higher at 24 h. Y7 and Y33 had significantly lower AEP production (especially at 12 h) and cell growth than Y3. Y2 and Y34 had comparable or higher AEP production at 24 h, but less than half at 12 h. Based on its high AEP productivity, Y3 was chosen for further study.

Table 1. Screening XPR2 multi-copy transformants for growth and AEP production during AEP induction
StrainTime (h)Cell densitya (Klett units)RatiobAEP Production (U mL−1)/100 KUaRatiob
  1. a

    Values presented as mean (SD). n is number of replicates.

  2. b

    Compared with W29-Δura3 values for samples taken at the same time.

W29-Δura312925 (136) n = 41.027 (2.2) n = 41.0
Y212525 (106) n = 20.5778 (25) n = 22.9
Y312540 (60) n = 30.58161 (40) n = 45.9
Y712480 n = 10.5267 n = 12.5
Y3312480 n = 10.5263 n = 12.3
Y3412503 (126) n = 30.5476 (12) n = 32.8
Y4412720 n = 10.79167 n = 16.1
W29-Δura3241488 (13) n = 41.023 (3.6) n = 41.0
Y224713 (113) n = 30.48291 (55) n = 312.6
Y324750 (105) n = 40.50286 (25) n = 412.4
Y724560 n = 10.38211 n = 19.2
Y3324600 n = 10.4072 n = 13.1
Y3424636 (164) n = 40.43333 (88) n = 414.5
Y44241020 n = 10.69212 n = 19.2

For Y44, AEP production was comparable to Y3 at 12 h and somewhat lower at 24 h. However, its growth was the least affected of any of the transformants producing high levels of AEP. A strain that significantly overproduced AEP but grew at near wild-type rates would be of interest because some changes in gene expression after AEP induction may not be specific for secretion stress but owing to changes in growth rate. Six additional transformants (strains Y1, Y5, Y22, Y38, Y50, and Y85) that performed almost as well as Y44 on plate assays were screened in AEP induction experiments. None grew as well as Y44 after AEP induction. At 12 h, AEP production was higher for Y44 than for six of the seven strains – Y85 was 20% higher but cell density was only 52%. Based on its better growth and its high AEP production at 12 h, Y44 was also chosen for further study.

Growth and AEP production of W29 Δura3, Y3, and Y44 during AEP induction

Each strain was transferred to four flasks containing AEP induction medium. Changes with time for cell density, AEP levels, and AEP levels/cell density for W29 Δura3, Y3, and Y44 are presented in Fig. 2. For the first 2 h, the lag in cell growth after transfer to induction medium was fairly similar. From 2 h to 3 h, the growth rates of strains Y3 and Y44 slowed compared with W29 Δura3 (Fig. 2a). After 3 h, their growth rates increased but remained lower than for W29 Δura3. The doubling time for W29 Δura3 from 3 to 6 h was 3.5 ± 0.1 h. From 3 h to 7h, Y44 (td of 4.1 ± 0.2 h) grew more slowly than W29 Δura3 and more rapidly than Y3 (td of 8.7 ± 0.7 h).

image

Figure 2. Growth and AEP production after transfer to AEP induction medium for W29 Δura3 (▲), Y3 (○), and Y44 (●). Data points represent mean ± SD of measurements from four independent cultures. (a) Growth – cell density vs. time, (b) AEP concentration vs. time, and (c) AEP concentration divided by cell density vs. time.

Download figure to PowerPoint

Over the first 3 h, AEP concentrations (AEP mL−1) were slightly higher for Y3 than for Y44 and both were substantially higher than for W29 Δura3 (Fig. 2b). After 3 h, the rate of increase for AEP concentration for Y3 slowed relative to Y44. From 4 h to 10 h, AEP concentrations for Y3 and W29 Δura3 were comparable.

Over the first 3h, AEP concentration per cell density values (Fig. 2c) for Y3 are slightly higher than for Y44 and both are higher than for W29 Δura3. Afterward, values for Y44 increase more rapidly than for Y3, and Y3 values are clearly higher than W29 Δura3 values.

Results for a similar AEP induction experiment used to produce samples for a preliminary gene expression study are shown for Y3 and for Y44 in Fig. 3. Data for all time points for both strains are presented in Table S3. As in Fig. 2, growth of Y3 started to slow compared with Y44 by 3 h, and AEP concentration was slightly higher for Y3 over the first 3 h. Afterward, AEP concentration increased much more rapidly for Y44.

image

Figure 3. Growth and AEP production after transfer to AEP induction medium for Y3 (○) and Y44 (●) cultures used as the sources of samples for a preliminary DNA microarray experiment. Only first 12 h of Y3 are shown. Data points represent mean ± SD of measurements from four independent cultures. Flask used as source of cells and the time when the sample was taken are indicated. (a) Growth – cell density vs. time, and (b) AEP concentration vs. time.

Download figure to PowerPoint

One way to compare AEP productivity of strains is to calculate differential rates. The differential rate of AEP production ((ΔAEP-related protein/mL/Δt)/(Δcell protein/mL/Δt)) was 41.4 ± 9.1% (n = 4) from 16 h to 28 h for Y3 and 26.0 ± 4.7% (n = 4) from 8 h to 10 h for Y44 (See Appendix S1 for calculation details). These results suggest that AEP-related production can become a high percentage of total protein production.

Differential rates can be misleading with regard to productivity, especially at very low growth rates. A high differential rate may be primarily owing to a low rate of increase in cell protein, and the absolute amount of AEP produced per cell mass over a given time period could be quite modest. A better measure of efficiency of AEP production is cell productivity (qp). It represents the increase in AEP concentration vs. the integral of viable cell density (cells*time/volume) (IVCD) – ΔAEP (U/L)/ΔIVCD (cells*h/L) – over the cumulative time interval (Adams et al., 2007). qp is the increase of concentration of AEP produced divided by the product of cell concentrations at different times multiplied by the time intervals during which AEP is being produced, that is, a rough estimate of the integral of cells*time/volume. For W29 Δura3, qp was 0.015 ± 0.001 mg (109 cells*h)−1 at 3 h and nearly constant between 4 h and 8 h – 0.037 ± 0.001 mg (109 cells*h)−1 (calculated from data in Fig. 2).

For Y3, the slope of a plot of AEP concentration vs. IVCD for flasks A, B, and D abruptly levels at about 15 mg L−1 of AEP, corresponding to 3 h when IVCD was 170–180 × 109 cells*h L−1 (Fig. S1). For flask C, the slope also decreases at this AEP level and time but not as severely. After 4 h, the slope increases in all flasks but is not as steep as from 2 to 3 h. A plot of qp vs. time shows the changes in cell productivity more directly (Fig. 4a). qp reaches a maximum of 0.069 ± 0.001 mg (109 cells*h)−1 at 3 h (4.6× the W29 Δura3 value at 3 h). From 3 to 4 h, AEP concentration was fairly constant but because IVCD continued to increase (Fig. S1) qp decreased. It reached 0.053 ± 0.005 mg (109 cells*h)−1 at 4 h, and then stabilized or increased slightly afterward (maximum value 0.059 ± 0.005 mg (109 cells*h)−1. qp of 0.069 mg (109 cells*h)−1 is equivalent to 1.66 pg cell−1 day−1 or 22 800 molecules cell−1 min−1 (See Appendix S1).

image

Figure 4. qp (AEP productivity (mg (109 cells h)−1) vs. time after transfer to AEP induction medium, calculated from the slope over the cumulative time interval of concentration of secreted AEP (mg mL−1) vs. integral of viable cell density (cells*time/volume) (IVCD) (see Fig. S1 and Fig. S2). Data points represent mean ± SD of measurements from four independent cultures. (a) Strain Y3 (○) and (b) strain Y44 (●).

Download figure to PowerPoint

For Y44, the slope of a plot of AEP concentration vs. IVCD does not level off at 3 h but is fairly constant from 3 h to 6 h and then it becomes less steep (Fig. S2). For Y44, the increase in qp was not as rapid as for Y3; qp was 0.052 ± 0.008 mg (109 cells*h)−1 at 3 h (3.5× the W29 Δura3 value at 3 h) (Fig. 4b). There was no decrease in qp between 3 h and 4 h and at 6 h qp reached 0.114 mg (109 cells*h)−1. Afterward, qp appeared to plateau. The maximum value attained was 0.118 ± 0.011 mg (109 cells*h)−1 at 9 h. So Y44 had a different pattern of cell productivity vs. time than W29 Δura3 and Y3, and ultimately Y44 was more productive than Y3. qp of 0.118 mg (109 cells*h)−1 is equivalent to 2.83 pg per cell day−1 or 38 800 molecules per cell min−1 (See Appendix S1).

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

In this manuscript, improvements in a system for rapidly increasing secretory pathway cargo load by AEP induction in Y. lipolytica are described. The system includes several important advances over the previously described system (Le Dall et al., 1994). In the previous system, highly inbred E129-related strains with 25–60 integrated copies of XPR2 grew rapidly for at least 10 h after transfer to a glucose/proteose peptone medium at pH 6.8 before AEP production began and then growth rate decreased for several hours before increasing again (Le Dall et al., 1994). A complication was that by then cells had lost copies of XPR2. With the improved strains and induction conditions described in this study, more AEP was produced more rapidly. Use of W29 Δura3-derived XPR2 multi-copy strains instead of E129-derived and 23°C instead of 28°C for AEP induction significantly increased AEP productivity. Removal of most of the low-molecular-weight N sources in proteose peptone by dialysis allowed more rapid AEP induction. Earlier and stronger AEP induction results in much earlier slowing of growth (by 2 to 3 h after transfer to AEP induction medium). Hardly any growth had occurred by this time and little or no of loss of XPR2 gene copies should be possible in this brief time period. This should minimize the complication found in the previous system of loss of copies of XPR2.

Another practical advantage with the new system is that AEP induction by 1 h and slowing of growth by 2 to 3 h vs. AEP induction after at least 10 h in the previous system narrows the window during which secretory pathway cargo loads are rapidly increasing. This should facilitate finding appropriate times for taking samples for kinetic analysis, which can be problematic (see below). Preliminary DNA microarray gene expression studies provided useful information about when to take the control samples and how often to take samples (see below). Also, the W29 Δura3 XPR2 multi-copy strains grow significantly more rapidly in defined medium with NH4+ as N source at pH 6.8 (used as preinduction medium in which AEP production is repressed) than do the E129-derived XPR2 multi-copy strains (precultured in YPD 5.0) making it possible to avoid pH changes upon transfer to AEP induction medium.

A very important advantage of using W29-derived vs. E129-derived strains is that W29-derived strains should be more ‘wild type’ for secretory pathway components and secretory pathway regulation. W29 is essentially a wild-type strain with a deletion of URA3 integrated into its genome. E129 is highly inbred; selection was for characteristics useful for doing classical genetics such as sporulation frequency and spore viability. This inbreeding is probably responsible for N-metabolism difficulties of E129-derived strains. The facts that W29-derived strains secreted much higher levels of AEP (this study) and heterologous proteins (Dalboge, 1997) compared with E129-derived strains suggest that the extensive inbreeding has also adversely affected secretory abilities of E129-derived strains. Therefore, W29-derived strains should be a much better choice for gene expression studies of secretion stress.

Additionally, gene expression studies of AEP induction should provide insights into how cells adapt to conditions where growth is dependent on uptake and metabolism of products of proteolysis by a secreted protease, in this case AEP. And again W29-derived strains with more ‘wild-type’ N-metabolism are clearly a better choice for such studies than E129-derived strains that grow very poorly on NH4+.

Many studies of effects of over expression of secreted proteins involve heterologous proteins. The fact that AEP is a homologous protein may offer advantages in terms of compatibility with secretory pathway and protein processing components, and this might contribute to the high rates of AEP production and secretion obtained.

Multi-copy XPR2 transformants constructed in W29 Δura3 and CX161-1B Δura3 were screened on plates and in AEP induction experiments. High AEP-producing strains Y3 and Y44, which differ in severity of growth defects after AEP induction, were characterized in more detail.

The reasons for (1) the more severe effects of AEP induction on the growth rate of Y3; (2) the decrease of qp after 3 h for Y3 vs. plateauing of qp after 6 h for Y44; and (3) the higher productivity of Y44 are not clear. Higher qp values for Y3 at early times indicate higher initial rates of AEP production. Most likely, Y3 has more integrated copies of XPR2. However, differences in integration site(s) may also affect responses. A lower XPR2 copy number in Y44 might be closer to the optimum number for maximum AEP productivity under these AEP induction conditions. Possibly around 3 h, the AEP cargo load overwhelmed the secretory capacity of Y3 resulting in decreased AEP secretion until the cells adapted to some extent. The later plateauing of qp for Y44 (6 h vs. 3 h), the higher growth rate and maximum qp could be explained by Y44 cells being more able to adapt to the lower level of secretion stress caused by lower rates of AEP production. The fact that qp for Y3 never recovers above levels reached at 3 h suggests that continuing higher levels of secretion stress compromise how efficiently it ultimately can adapt.

In S. cerevisiae, expression of over 25% of genes is linearly correlated with growth rate (Brauer et al., 2008). AEP induction in XPR2 multi-copy strains slows growth. Therefore, many of the changes in gene expression that would be detected during AEP induction may not be a direct response to secretory cargo overload but to slowing of growth. Use of continuous cultures (where growth rates can be controlled) to compare strains or responses to environmental conditions overcomes this complication. However, with the Y. lipolytica XPR2 multi-copy strains, this would not be an attractive option because XPR2 gene copies would most likely be lost during initial stabilization of the continuous cultures.

Alkaline extracellular protease induction with XPR2 multi-copy strains should provide an excellent system for following gene expression changes during acute increases in secretory pathway cargo load. In preliminary studies using DNA microarrays, gene expression changes for Y3 and Y44 (sampling times are indicated in Fig. 3) during AEP induction were examined. Data for at least three time points were obtained for over 5900 genes for Y3 and for over 5200 genes for Y44 (data not shown). After transfer to AEP induction medium, control samples were taken before growth slowed (at 1.5 h for Y3 and at 2 h for Y44). The next samples were taken after growth had already begun to slow (at 4 h for both strains). A major conclusion of these studies that should be useful for future studies was that earlier control samples and more closely spaced time points must be examined to detect acute responses to secretion stress. Two results support this conclusion. First, there was little increase in XPR2 and KAR2 transcript levels at 4 h, and the fluorescent intensity for these genes in the control samples was among the highest found for any of the genes. This suggests that XPR2 and KAR2 were already almost fully induced in the control samples. Second, repression but not induction of genes involved in protein translocation, protein folding in the ER or protein transport into and through the secretory pathway was found at 4 h. Induction of some of these genes had been expected, but only finding repression is consistent with down-tuning of expression of genes that had been already significantly induced in the control samples.

The YlIRE1 (Babour et al., 2008) and YlHAC11 (Oh et al., 2010) orthologs have been identified and roles in UPR demonstrated, indicating that Y. lipolytica has a version of UPR. Whether or not UPR is induced during AEP induction is unclear. At 4 h during AEP induction, only three (CSR1, PMT2, and OPI3) of the 103 genes (categorized as being involved in protein secretion or in biogenesis of secretory organelles) that were found to be up-regulated during UPR in S. cerevisiae by Travers et al. (2000) were up-regulated (data not shown). In Y. lipolytica, the almost full induction of KAR2, often used as a reporter for UPR regulation, in the control samples suggests that evidence for UPR-like response might have been missed if responses to increased secretory pathway cargo load had already begun when the control samples were taken.

In conclusion, these new Y. lipolytica XPR2 multi-copy strains and improved AEP induction conditions, in combination with DNA microarray gene expression analysis, should provide a powerful system for kinetic analysis of responses to rapid increases in secretory pathway cargo load.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Partial support for D.M.O. was from a UC Davis Faculty Research Grant. J.M.N. was supported by CNRS and INRA. We thank L. Oldford for preliminary induction experiments, E. Dedeoglu for determining temperature effects on growth and AEP production, and M-T. LeDall for advice on E129-derived strains. We especially thank C. Gaillardin for helpful suggestions for improving the manuscript.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information
  • Adams D, Korke R & Hu WS (2007) Application of stoichiometric and kinetic analyses to characterize cell growth and product formation. Methods in Biotechnology, Vol. 24: Animal Cell Biotechnology: Methods and Protocols, 2nd edn (Portner R, ed.), pp. 269284. Humana Press, Inc., Totowa, NJ.
  • Arvas M, Pakula T, Lanthaler K, Saloheimo M, Valkonen M, Suortti T, Robson G & Penttila M (2006) Common features and interesting differences in transcriptional responses to secretion stress in the fungi Trichoderma reesei and Saccharomyces cerevisiae. BMC Genomics 7: 32.
  • Babour A, Kabani M, Boisrame A & Beckerich JM (2008) Characterization of Ire1 in the yeast Yarrowia lipolytica reveals an important role for the Sls1 nucleotide exchange factor in unfolded protein response regulation. Curr Genet 53: 337346.
  • Barth G & Gaillardin C (1996) Yarrowia lipolytica. Non-conventional Yeasts in Biotechnology. A Handbook (Wolf K, ed.), pp. 313388. Springer-Verlag, Berlin.
  • Brauer MJ, Huttenhower C, Airoldi EM, Rosenstein R, Matese JC, Gresham D, Boer VM, Troyanskaya OG & Botstein D (2008) Coordination of growth rate, cell cycle, stress response, and metabolic activity in yeast. Mol Biol Cell 19: 352367.
  • Casagrande R, Stern P, Diehn M et al. (2000) Degradation of proteins from the ER of S. cerevisiae requires an intact unfolded protein response pathway. Mol Cell 5: 729735.
  • Dalboge H (1997) Expression cloning of fungal enzyme genes; a novel approach for efficient isolation of enzyme genes of industrial relevance. FEMS Microbiol Rev 21: 2942.
  • Davidow LS, Apostolakos D, O'Donnell MM, Proctor AR, Ogrydziak DM, Wing RA, Stasko I & DeZeeuw JR (1985) Integrative transformation of the yeast Yarrowia lipolytica. Curr Genet 10: 3948.
  • Enderlin CS & Ogrydziak DM (1994) Cloning, nucleotide sequence and functions of XPR6, which codes for a dibasic processing endoprotease from the yeast Yarrowia lipolytica. Yeast 10: 6779.
  • Fox RM, Hanlon CD & Andrew DJ (2010) The CrebA/Creb3-like transcription factors are major and direct regulators of secretory capacity. J Cell Biol 191: 479492.
  • Gaillardin CM, Charoy V & Heslot H (1973) A study of copulation, sporulation and meiotic segregation in Candida lipolytica. Arch Mikrobiol 92: 6983.
  • Gasser B, Maurer M, Gach J, Kunert R & Mattanovich D (2006) Engineering of Pichia pastoris for improved production of antibody fragments. Biotechnol Bioeng 94: 353361.
  • Gasser B, Sauer M, Maurer M, Stadlmayr G & Mattanovich D (2007) Transcriptomics-based identification of novel factors enhancing heterologous protein secretion in yeasts. Appl Environ Mictobiol 20: 64996507.
  • Guillemette T, van Peij NN, Goosen T et al. (2007) Genomic analysis of the secretion stress response in the enzyme-producing cell factory Aspergillus niger. BMC Genomics 8: 158.
  • Jorgensen TR, Goosen T, Hondel CA, Ram AF & Iversen JJ (2009) Transcriptomic comparison of Aspergillus niger growing on two different sugars reveals coordinated regulation of the secretory pathway. BMC Genomics 10: 44.
  • Le Dall MT, Nicaud JM & Gaillardin C (1994) Multiple-copy integration in the yeast Yarrowia lipolytica. Curr Genet 26: 3844.
  • Madzak C, Gaillardin C & Beckerich JM (2004) Heterologous protein expression and secretion in the non-conventional yeast Yarrowia lipolytica: a review. J Biotechnol 109: 6381.
  • Martini A (1993) Origin and domestication of the wine yeast Saccharomyces cerevisiae. J Wine Res 4: 165176.
  • Matoba S, Morano KA, Klionsky DJ, Kim K & Ogrydziak DM (1997) Dipeptidyl aminopeptidase processing and biosynthesis of alkaline extracellular protease from Yarrowia lipolytica. Microbiology 143(Pt 10): 32633272.
  • Ogrydziak DM & Scharf SJ (1982) Alkaline extracellular protease produced by Saccharomycopsis lipolytica CX161-1B. J Gen Microbiol 128: 12251234.
  • Ogrydziak DM, Demain AL & Tannenbaum SR (1977) Regulation of extracellular protease production in Candida lipolytica. Biochim Biophys Acta 497: 525538.
  • Oh MH, Cheon SA, Kang HA & Kim JY (2010) Functional characterization of the unconventional splicing of Yarrowia lipolytica HAC1 mRNA induced by unfolded protein response. Yeast 27: 443452.
  • van Anken E, Romijn EP, Maggioni C, Mezghrani A, Sitia R, Braakman I & Heck AJ (2003) Sequential waves of functionally related proteins are expressed when B cells prepare for antibody secretion. Immunity 18: 243253.
  • van Ooyen AJ, Dekker P, Huang M, Olsthoorn MM, Jacobs DI, Colussi PA & Taron CH (2006) Heterologous protein production in the yeast Kluyveromyces lactis. FEMS Yeast Res 6: 381392.
  • Patil C & Walter P (2001) Intracellular signaling from the endoplasmic reticulum to the nucleus: the unfolded protein response in yeast and mammals. Curr Opin Cell Biol 13: 349355.
  • Peng R-W & Fussenegger M (2009) Engineering the secretory pathway in mammalian cells. Cell Engineering, Vol. 6 (Al-Rubeai M, ed.), pp. 233248. Springer Science+Business Media B.V., Dordrecht.
  • Pulvirenti T, Giannotta M, Capestrano M et al. (2008) A traffic-activated Golgi-based signalling circuit coordinates the secretory pathway. Nat Cell Biol 10: 912922.
  • Ron D & Walter P (2007) Signal integration in the endoplasmic reticulum unfolded protein response. Nat Rev Mol Cell Biol 8: 519529.
  • Sambrook J, Fritsch EF & Maniatis T (1989) Molecular Cloning: A Laboratory Manual, 2nd edn. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.
  • Schuck S, Prinz WA, Thorn KS, Voss C & Walter P (2009) Membrane expansion alleviates endoplasmic reticulum stress independently of the unfolded protein response. J Cell Biol 187: 525536.
  • Sharp PM & Cowe E (1991) Synonymous codon usage in Saccharomyces cerevisiae. Yeast 7: 657678.
  • Sharp PM & Li WH (1987) The codon adaptation index – a measure of directional synonymous codon usage bias, and its potential applications. Nucleic Acids Res 15: 12811295.
  • Travers KJ, Patil CK, Wodicka L, Lockhart DJ, Weissman JS & Walter P (2000) Functional and genomic analyses reveal an essential coordination between the unfolded protein response and ER-associated degradation. Cell 101: 249258.

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information
FilenameFormatSizeDescription
fyr846-sup-0001-Fig.S1.xlsapplication/msexcel85KFig. S1. AEP concentration vs. IVCD for Y3 (○) during AEP induction. (a) flask A, (b) flask B, (c) flask C, and (d) flask D.
fyr846-sup-0002-Fig.S2.xlsapplication/msexcel85KFig. S2. AEP concentration vs. IVCD for strain Y44 (●) during AEP induction. (a) flask A, (b) flask B, (c) flask C, and (d) flask D.
fyr846-sup-0003-TableS1.xlsapplication/msexcel37KTable S1. Codon adaptation index of select Saccharomyces cerevisiae and Yarrowia lipolytica genes involved in the secretory pathway.
fyr846-sup-0004-TableS2.docWord document168KTable S2. Plate screening for AEP high producing strains.
fyr846-sup-0005-TableS3.xlsapplication/msexcel35KTable S3. Cell density and AEP concentration data for all time points for Y3 and Y44 samples used in a preliminary DNA microarray experiment.
fyr846-sup-0006-AppendixS1.docWord document26KAppendix S1. Calculations of AEP-related production as percentage of total protein production, of qp and of AEP molecules secreted per cell min−1.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.