Biotechnology Letters Received ist August.

Vol II

No 9

609-614

CONTINUOUS CULTURE OF ESCHERICHIA COLI FOR EXTRACELLULAR PRODUCTION OF RECOMBINANT AMYLASE Peter Alexander*' and Patrick J. Oriel Department of Microbiology and Public Health and David A. Glassner and Eric A. Gmlke Dep~at-~ent of Chemical Engineering Michigan State University East Lansing, MI 48824 SUMMARY

A recombinant Escherichia coli, grown in continuous culan'e, expressed a Bacillus stearothermophilus tx-amylase at 100-fold higher activities than the B. stearothermophilus itself. Excretion of the ixamylase to the supernatant was shown and found to be independent of the growth rate of the organism. Eleven to eighteen percent of the ct-amylase was found in the supernatant~ Dilution rotes, or cell growth rates, ranging from 0.1 to 1.0 hourst were shown not to affect the compartmentation of the amylase and B-galactosidase. INTRODUCTION Recombinant plasmid systems for overproduction o f the proteins have been developed, but are frequently limited by the fact that product is contained in the cytoplasm or periplasm. The excretion of recombinant proteins from the well understood organism, Escherichia coli would be highly desirable in order to facilitate easier recovery and purification, and avoid proteolysis of the desh-ed protein. Several promising systems for recombinant protein excretion by E. coli hm~e recently been described (Abramsen et. al., 1986; Kato et. al., 1987; Georgiou et. al., 1988).

These protein

excretions seem to occur only at slow growth rates. Knowledge of the growth rate dependence in recombinant systems will aid in understanding the mechanism of release and optimizing the protein release. An excretion system for a-amylase was described by Oriel and Schwacha (1988). In batch cultures of the recombinant microbe, up to 25 percent of the a-amylase activity was measured in the culture supernatant.

To better understand the mechanism of the protein release and to help optimize the

protein release, a study of this recombinant organism was begun. This work examines the effect of growth rate on the production and compartmentation of the recombinant protein.

, Present Address: Seragen, Inc., 97 South St., Hopkinton, MA 01748

609

(1989)

MATERIALS AND METHODS Strains, Plasmids and Culture Conditions Eschericlu'a coli strain DH1 (Hanahan, 1983) containing the tx-ainylase gene from Bacillus stearothermophilus (ATCC 29609) inserted into the pBR322 plasmid was designated EC198 and used in the experiments described. Cultures were grown in LB broth (GIBCO) containing 100 ~tg/ml ampicillin (Sigma Chemical) in an airlift fermentor (LH Fermentation) with a 2.2 1 working volume as described by Grulke and Glassner (1988). The aeration rate was 150 cm ~ per minute.

Cell Fractionations For determining the compartmentation of proteins, 10 ml samples were collected and the cell density was measured using a Klett-Summerson colorimeter (580-620 nm filter). The cells were collected by centrifugation at 12,000 g for 10 minutes at 4~ The supernatant was filtered using a 0.2 I.tm filter and the filtrate was stored on ice until assayed. This fraction represents the extracellular (excreted) fraction. The cell pellet was washed in 10 mM Tris HC1, pH 7.5 and the cells were subjected to osmotic shock as described by Nossal and Hepple 1966). The resulting periplasmic fraction was filtered through a 0.2 I.tm filter and stored on ice until assayed. The shocked cell pellet was washed with 50 mM Tris HC1, pH 7.5 and the cells lysed with 4 mg/ml lysozyme contained in the buffer at room temperature (22-25~ The lysate resulting from the lysozyme treatment was stored on ice as the cytoplasmic fraction and was assayed with the other fractions. The cytoplasmic fraction contains soluble as well as membrane associated components.

Enzyme Assays or-amylase activity was assayed at 70~ as described (Oriel and Schwacha, 1988) using soluble starch (Difco) as the substrate, l]-galactosidase production was induced by the addition of 0.5 mM isopropyl-D-galactoside (IgrG) and was assayed at 37~ as described (Miller, 1972) using onitrophenyl-I]-g~actoside (o-NPG) as the substrate.

RESULTS AND DISCUSSION Production of recombinant amylase by continuous culture fermentation in an airlift reactor was examined using E. coli.

Figure 1 shows the a-amylase activity at various continuous culture

dilution rates. Total amylase activity (units/ml) was highest at the lowest dilution rate tested. This result is not surprising since the cell density is highest at the lowest dilution rate. Under these conditions, amylase expression by the recombinant E. coli EC198 reached 230 units/ml at a dilution rate of 0.1 hours -1. Table 1 shows the steady-state cell density and cell productivity for each of the dilution rates.

610

"

I

I

i

I

I

I

I

200

160

~>,~

120

I

I

i Total

.>, 3~----

I

Supernatant Periplasm [] C y t o p l a s m

-

Figure 1 Alpha-Amylase Activity

_

8o 4O I

~'

i~-~---

0.2

I,

0.4

m

q

0.6

Dilution Rate

1

0.8

~-1,0

(hours "1 ) Table 1.

Cell Density and Cell Productivity Cell Density (Kletts)

Dilution Rate (1/hr) 0.1 0.2 0.35 0.5 1.0

Cell Productivity (Kletts/hr)

345 289 206 185 94

34.5 57.8 72.1 92.5 94.0

Figure 2 shows that specific amylase activity (activity/cell) also increases with decx:easing dilution rate, suggesting that the increase is not solely the result of increased cell density in the culture. This i

i

i

[

'

i

i

e

0.6

>

=

i

i

Total

o Supernatant Periplasm [] C y t o p l a s m

F i g u r e 2. Specific A l p h a - A m y l a s e Activity

0.4

o r0.2 D m

Q. (n

~ ,

0.2

0.4 Dilution

0

~

I).6 0'.8 Rate (hours"1 )

,

1.0

difference may be related to changes in plasmid copy number with growth rate.

Lin-Chao and

Bremer (1986) found that total plasmid content of the E. coli Culture decreased with increasing growth rate. The highest amylase activity expressed by B. stearothermophilus (ATCC 29609) was less than 5 units/ml (Glassner et al., 1989). Thus, the EC198 recombinant system increased the amylase concentration in the culture by two orders of magnitude and released 10-fold more extracellular amylase than the B. stearothermophilus strain expressing the same gene. However, most of the amylase produced is retained in the periplasm rather than excreted.

611

It was of interest to examine the compartmentation of enzymes with growth rate. Compartmentation of the amylase by EC198 was not growth rate dependent as the a-amylase activity in each compartment increased proportionately as the dilution rate decreased.

The results shown in Figure

1 indicate that 11 to 18 percent of the amylase is excreted into the culture medium (18 percent at a dilution rate of 0.1 hour-1 corresponds to 40 units/ml). The periplasm contained 61 to 79 percent of the activity and 5 to 26 percent remained in the Cytoplasmic fraction. The integrity of cellular compartments was examined in order to discriminate between excretion and non-specific release due to ceil lysis. ]3-gaiactosidase is an inducible enzyme normally found in the E. coli cytoplasm. Analysis of fractionated EC198 cells induced during continuous culture showed

that l~-galactosidase remains with the cytoplasmic compartment at each dilution rate tested (Table 2). Further, the ratio of extracellular to periplasmic amylase remained constant at each dilution rate (Table 3). Table 2. 13-galactosidase Compartmentation at Several Dilution Rates g-galactosidase Activity (BGA) (Units/ml) Dilution Rate (1/hr)

1.0 BGA

0.2

0.1

~

BGA

%

BGA

%

Supematant Periplasm Cytoplasm

20 38 508

3.5 6.7 89.8

79 95 4329

1.8 2.1 95,6

145 119 5084

2,7 2.2 95.0

Total

566

100.0

4503

'99.5

5348

99.9

Table 3. Ratio of Supernatant to Periplasmic Enzyme Activity at Various Dilution Rates Dilution Rate

Ratio

1,00 0.50 0.35 0.20 0.10

0.22 0.20 0.14 0.25 0.27

Thus, it was concluded that cell envelope integrity was maintained at all dilution rates and that the extracellular release of amylase was not due to membrane leakiness or cell lysis.

612

Figure 3 shows the specific or-amylase productivity plotted versus several dilution rates. The specific productivity (units/hr-klett) is defined as the amylase activity (unitstl-hr) divided by the ceil density (klett) and multiplied by the total volume (1). Specific amylase productivity (productivity/cell) in each cellular compartment increased with increasing dilution rate (Figure 3) and suggests that specific productivity is highest in rapidly growing cells. This effect is probably the result of increased

--

320 280

i 240 2OO 160 120 o 80

9 Total o Supernatant 9 Periplasm

~

j.t-

0'.6

I

, m

"~ r

40 i

024

0.2

().8 Dilution Rate (hours-1)

1.

I0

Figure 3. Specfic Alpha-Amylase Activity

plasmid copy number in EC198 ceils when grown under optimal growth conditions.- Lin-Chao and Bremer (1986) observed an increase in pBR322 copy number per cell as growth rate increased from 0.6 to 2.5 doublings/hr. In contrast, Table 4 shows that volumetric amylase productivity decreased slightly with increasing dilution rate. In Table 1, we show that volumetric cell productivity increases with increasing dilution rate. This indicates that the best productivity and amylase concentration were obtained at the lowest Table 4. Volumetric Productivity of et-Amylase by Cell Compartment Dilution Rate (1/hr)

1.0

Cell Compartment

0.5

0.35

0.2

0.1

a-Amylase Productivity (x 10.3 units/hr)

Supematant Periplasm Cytoplasm

2.7 12.2 1.1

2.3 11.5 5.0

2.1 14.9 1.8

3.8 15.3 1.1

3.8 14.3 5.0

Total

16.0

18.8

18.8

20.2

23.1

613

dilution rate tested (0.1 hour-l). These results appear to be quite different than those reported by Seo and Bailey (1986). They found sharp optima for both activity and volumetric productivity in E. coli expressing 13-1actamase in continuous culture, when dilution rate was varied between 0.2 and 0.7 hour 1. Cell growth and metabolism for the results presented here may have been oxygen limited. We have shown that recombinant E. coli grown in continuous culture can express a B. stearothermophilus a-amylase at 100-fold higher activities than B. stearothermophilus itself, and

released 10-fold more of the activity to the culture medium. The following are key conclusions of the study: i) compartmentation of the enzyme was unaffected by dilution rate; The a-amylase in the supernatant was probably not due to cell breakage; ii) the specific productivity increased with dilution rate; This needs further investigation to determine the cause, and iii) excretion to the supernatant by fast growing cells has been shown and is an improvement over the use of leaky cells which exhibit poor growth (Anderson, et. al., 1979; Lazzaroni and Portailer, 1981). ACKNOWLEDGEMENT

This work was supported by the Michigan Biotechnology Institute. REFERENCES

Abramsen, L., Moks, T., Nilsson, B., and Uhlen, M. (1986). Nucl. Ac. Res. 14, 7487-7500. Anderson, J.J., Wilson, J.M., and Oxender, D.L. (1979). J. Bacteriol. 140, 351-358. Georgiou, G., Shuler, M.L., and Wilson, D.B. (1988). Biotechnol. Bioeng. 32, 741-748. Glassner, D.A., Grulke, E.A., and Oriel, P.J. (1989). Biotechnol. Prog. 5, 31-39. Hanahan, D. (1983). J. Mol. Biol. 166, 557-580. Kato, C., Kobayashi, T., Kudo, T., Furusato, T., Murakami, Y., Tanaka, T., Baba, H., Oishi, T., Ohtsuka, E., Ikehara, M., Tanagida, T., Kato, H., Moriyama, S., and Horikoshi, K. (1987). Gene 54, 197-202. Lazzaroni, J.-C., and Portalier (1981). J. Bacteriol. 145, 1351-1358. Lin-Chao, S., and Bremer, H. (1986). Mol. Gen. Genet. 203, 143-149. Miller, J.H. (1972). Experiments in Molecular Genetics, Cold Spring Harbor, New York: Cold Spring Harbor Laboratories. Nossal, N.G., and I-Iepple, L.A. (1966). J. Bio. Chem. 241, 3055-3062. Oriel, P.J., and Schwacha, A. (1988). Enz. Microb. Technol. 10, 41-46. Seo, J-H., and Bailey, J.E. (1986). Biotechnol. Bioeng. 28, 1590-1594.

614