Detection of L-Lactate In Polyethylene Glycol Solutions Confirms Identity of Active Site Ligand in Proline Dehydrogenase Structure

Abstract

Polyethylene glycol (PEG) is often used in protein crystallography as a low ionic strength precipitant for crystallization and a cryoprotectant for low temperature data collection. Prompted by the discovery of an apparent L-lactate molecule bound in the active site of the E. coli PutA proline dehydrogenase domain crystal structure, we measured the L-lactate concentration of several PEG solutions. Fifty percent (w/v) solutions of PEGs with molecular weight 3000, 4000, and 8000 contain millimolar levels of L-lactate. In contrast, L-lactate was not detected in solutions of PEG monomethyl ethers or PEG 3350. These results help explain why L-lactate was present in the proline dehydrogenase domain crystal structure. This work also has implications for the crystallization of enzymes that bind L-lactate.

1. Introduction

Polyethylene glycol (PEG) was first used successfully as a precipitating agent for protein crystallization in the mid 1970s, when Ward et al. reported the structure of deoxyhemoglobin using crystals grown in PEG (Ward et al., 1975). The next year McPherson firmly established PEG as an important crystallization reagent by crystallizing 13 proteins using a variety of PEGs. He concluded that PEG might be the best initial trial reagent for crystal screening (McPherson, 1976). PEG monomethyl ether (PEGMME) was added to the protein crystallization arsenal in 1994 (Brzozowski and Tolley, 1994). Today, PEG and PEGMME are the most commonly used precipitating agents for protein crystallization, based on their inclusion in about half of the reagents in Crystal Screen, Wizard, and Index crystal screens. Therefore, it is of interest to know the identities and concentrations of contaminating species present in commercially available PEGs and PEGMMEs. For example, Ray, Jr. and Puvathingal discovered that rabbit muscle phosphoglucomutase lost activity within a few hours after exposure to 3 % PEG 400, which prompted them to develop a chromatographic procedure for removing contaminating aldehydes and peroxides from PEGs (Ray and Puvathingal, 1985).

Here, we present an analysis of L-lactate levels in PEG solutions commonly used in protein crystallization. This study was motivated by the discovery of an apparent L-lactate molecule bound in the active site of the E. coli PutA proline dehydrogenase (PRODH) domain crystal structure (Lee et al., 2003), despite the fact that lactate was not knowingly added to the protein solution used for crystallization (Nadaraia et al., 2001). Since the crystals were grown in 24 % (w/v) PEG 3000, and L-lactate had been detected in PEG 1000 previously (Pollegioni et al., 2002), we suspected that PEG might be the source of L-lactate. Using a coupled enzymatic assay, we find that stock solutions of PEG 3000, 4000, and 8000 contain millimolar levels of L-lactate, while PEGMME 2000, PEGMME 5000, and PEG 3350 are apparently free of L-lactate. These results confirm the identity and source of the active site ligand in the PRODH crystal structure. They also suggest that PEGMMEs and PEG 3350 might be useful reagents for crystallizing PRODH and other enzymes known to bind L-lactate.

2. Material and methods

The procedure described by Noll was used to determine L-lactate levels in aqueous PEG solutions (Noll, 1983). This method is an end point assay involving two enzymes, NAD-linked L-lactate dehydrogenase (LDH) and L-alanine aminotransferase (ALT). LDH catalyzes the oxidation of L-lactate to pyruvate, thereby reducing NAD⁺ to NADH. ALT converts pyruvate and L-glutamate to L-alanine and 2-oxoglutarate. The ALT reaction is necessary to remove pyruvate from the system because the equilibrium of the LDH reaction favors the formation of lactate from pyruvate. The final NADH concentration, which is monitored by the absorbance at λ=339 nm, is proportional to the L-lactate concentration in the sample. The detection limit of this method is 0.07 mM (Noll, 1983).

The assay protocol was used as described previously (Noll, 1983) except the concentrations of LDH, ALT, and NAD⁺ were increased 4x, 4x, and 5x, respectively, in order to decrease the overall reaction time (Noll, 1983). The increased concentrations of LDH, ALT, and NAD⁺ were within recommended ranges (Noll, 1983). Rabbit muscle LDH and porcine heart ALT were purchased from Sigma (catalogue numbers L1254 and G9880), and used without further purification. The following reagents were purchased from Sigma: L-(+)-lactate (catalogue number L1750), NAD⁺ (catalogue number N1511), and L-glutamic acid (catalogue number 49449). The assays were performed using a Cary 100 UV-visible spectrophotometer equipped with a 6-by-6 multi-cell transporter.

Several commercially available PEGs were tested for the presence of L-lactate. PEGs of various molecular weights were purchased from Fluka and Sigma (see Table 1 for catalogue numbers) and aqueous solutions were prepared by dissolving the PEGs in deionized water (18 MΩ) produced by a MilliQ Synthesis water purification system. Selected reagents from Crystal Screen (Hampton Research), Index (Hampton Research), and Wizard 1 (Emerald Biostructures) crystals screens were also analyzed. The L-lactate concentrations reported in Table 1 for PEG 3000 (Fluka), PEG 4000 (Fluka), and PEG 3350 (Sigma) are the averages of four, five, and two trials, respectively. The L-lactate concentrations reported for all other reagents are the results of single trials. Note, however, that there is replication in the data due to the analysis of PEGs from multiple sources. For example, PEG 8000 from five sources was tested (Fluka and four screen reagents) and the five measurements yielded similar results (Table 1).

Table 1.

L-lactate concentrations of PEG-containing reagents^†

Reagent	[PEG] (% w/v)	[L-lactate] (mM)
PEG 3000 (Fluka 81227)	50	3.5 ± 0.2
PEG 4000 (Fluka 81240)	50	3.0 ± 0.2
PEG 4000 (Crystal Screen #6)	30	2.1
PEG 8000 (Fluka 81268)	50	2.3
PEG 8000 (Crystal Screen #28)	30	1.8
PEG 8000 (Wizard 1 #1)	20	1.2
PEG 8000 (Wizard 1 #17)	30	1.8
PEG 8000 (Wizard 1 #31)	20	1.1
PEG 3350 (Sigma P-4338)	50	BD
PEG 3350 (Index #45)	25	BD
PEGMME 2000 (Fluka 81321)	50	BD
PEGMME 2000 (Index #47)	28	BD
PEGMME 5000 (Fluka 81323)	50	BD
PEGMME 5000 (Index #46)	20	BD

^†BD = below detection limit

3. Results and discussion

A standard calibration curve was constructed using L-lactate samples with concentrations of [L-lactate] = 0, 0.5, 1.0, 1.5, 2.0, 2.5 mM. The raw data for a typical standard curve are shown in Figure 1. The method employed is an endpoint assay in which L-lactate should be completely consumed. The endpoints in the present studies were typically reached within about 60 minutes, as shown in Figure 1. The resulting calibration curve (not shown) demonstrated that the method in our hands exhibits good precision and linearity up to [L-lactate] = 2.5 mM, and that L-lactate levels in the range 0.2 – 2.5 mM can be measured with confidence.

Fig. 1.

Enzymatic detection of L-lactate in solution. The curves show raw data from the analysis of six L-lactate standards ([L-lactate] = 0.0, 0.5, 1.0, 1.5, 2.0, 2.5 mM) and a 10 % (w/v) PEG 3000 stock solution.

Several PEG reagents were analyzed, including the PEG 3000 solution used in the crystallization, heavy atom soaking, and cryoprotection studies for the E. coli PutA PRODH domain (Table 1). The raw data for a 10 % (w/v) PEG 3000 solution (Fluka) are shown in Figure 1. The endpoint absorbance is 0.285, which corresponds to [L-lactate] = 0.67 mM. The PRODH domain was crystallized and cryoprotected in 24 % (w/v) PEG 3000, so the concentration of L-lactate in the crystals used for structure determination was approximately1.6 mM. This result helps explain the observation of a strong electron density feature shaped like an L-lactate molecule near the FAD cofactor in the PRODH structure (Lee et al., 2003). Given that L-lactate is a known competitive inhibitor of PRODH enzymes (Scarpulla and Soffer, 1978) the discovery of L-lactate in PEG 3000 confirms the identity and source of the active site ligand in the PRODH structure.

Since PEG 3000 contains significant L-lactate, it is unsuitable for crystallization of PRODH without inhibitors and PRODH complexes with other inhibitors. Therefore, other PEGs were analyzed in order to find a suitable precipitating agent for future crystallization experiments (Table 1). As with PEG 3000, millimolar levels of L-lactate were measured in 50 % (w/v) solutions of PEG 4000 (Fluka) and PEG 8000 (Fluka). Crystal screen reagents containing PEG 4000 and PEG 8000 also contain 1 – 2 mM L-lactate. These results are consistent with a previous measurement of 0.4 g/kg L-lactate in PEG 1000 (Pollegioni et al., 2002). Interestingly, L-lactate was not detected in a 50 % (w/v) solution of PEG 3350 (Sigma), nor in Index #45, which contains 25 % (w/v) PEG 3350. Likewise, the two PEGMMEs analyzed (MW 2000 and 5000) contain no detectable L-lactate.

These results underscore the fact that reagents used for protein purification and crystallization may contain significant levels of contaminating species. And, the degree of contamination may vary from vendor to vendor and product to product. For example, it is interesting that PEG 3000 and PEG 4000, both from Fluka, contain significant L-lactate, yet PEG 3350 from Sigma and PEGMMEs from Fluka appear to be free of significant L-lactate contamination. Perhaps differences in the manufacturing processes of these products account for the differences in L-lactate contamination. It is also not known whether the age of the PEG solution is an important variable. While the present study certainly is not exhaustive, it does suggest that caution should be exercised when using PEGs in the study of enzymes that bind L-lactate, such as PRODH, LDH, and D-amino acid oxidase (Umhau et al., 2000; Pollegioni et al., 2002). And, PEG 3350 and PEGMMEs should be useful precipitating agents in future crystallization studies of PRODH enzymes.

Synopsis.

Some commercially available PEGs in the molecular weight range 3000-8000 are contaminated with L-lactate. This result helps explain the discovery of L-lactate bound in the active site of the E. coli PutA proline dehydrogenase domain crystal structure.