Skip to main content
NCBI home page
Glycobiology logoLink to Glycobiology
. 2012 Jun 27;22(10):1302–1310. doi: 10.1093/glycob/cws102

Hyaluronan synthase polymerizing activity and control of product size are discrete enzyme functions that can be uncoupled by mutagenesis of conserved cysteines

Paul H Weigel 1,1, Bruce A Baggenstoss 1
PMCID: PMC3425326  PMID: 22745284

Abstract

Streptococcus equisimilis hyaluronan (HA) synthase (SeHAS) contains four cysteines (C226, C262, C281 and C367) that are conserved in the mammalian HAS family. Previous studies of single Cys-to-Ser and all possible Cys-to-Ala mutants of SeHAS found that: the Cys-null mutant is active, Cys modification inhibits HAS activity and the conserved cysteines are clustered at the membrane–enzyme interface in substrate-binding sites (Kumari K, Weigel PH. 2005. Identification of a membrane-localized cysteine cluster near the substrate binding sites of the Streptococcus equisimilis hyaluronan synthase. Glycobiology. 15:529–539). We re-examined these Cys mutants using a single technique (size exclusion chromatography–multi-angle laser light scattering) that allows simultaneous assays on the same sample for both HA synthesis activity and HA product size. Among 18 mutants compared with wild type, 4 showed no change in either function and 3 showed changes in both (decreased activity and HA size). Only one of the two functions was altered in 11 other mutants, which showed either decreased polymerizing activity or product size. No mutants made larger HA, 8 made smaller HA and 10 showed no change in HA size. Nine mutants showed no change in activity and nine were less active. The mutants fell into four of nine possible groups in terms of changes in HA size or synthesis rate (i.e. none, increased or decreased). Specific Cys residues were associated with each mutant group and the pattern of effects on both functions. Thus, the four conserved Cys residues, individually and in specific combinations, influence the rate of sugar assembly by HAS and HA product size, but their participation in one function is independent of the other.

Keywords: active-site cysteines, elongation rate, polymer length, size control, size control model

Introduction

Linear chains of hyaluronan (hyaluronic acid, HA) in vivo range in size up to 25,000 repeating disaccharide units (∼10 MDa), which contain N-acetyl-d-glucosamine (GlcNAc) and d-glucuronic acid (GlcUA). In vertebrates and streptococcal species such as Streptococcus equisimilis, HA is made by Class I HA synthases (HASs), which are lipid-dependent integral multiple membrane-domain proteins (Weigel and DeAngelis 2007). Distinct genes in mammals encode three HAS isozymes designated HAS1, HAS2 and HAS3. HA is also a major constituent of skin, vitreous humor in the eye and synovial fluid in joints. It has been known for >50 years that the biomechanical functions of very large MDa HA help in retaining fluids and tissue hydration, stabilizing cell-free spaces during development or wound healing and maintaining the structural integrity of differentiated tissues, such as cartilage. More recently, HA oligosaccharides were shown to be cell-signaling ligands, e.g. capable of stimulating angiogenesis or gene expression in macrophages (West et al. 1985; Noble et al. 1996), and intermediate-size HA (200–500 kDa) stimulates pro-inflammatory responses in a variety of pathologies (Stern et al. 2006). The recognition that there are different size dependencies for the multiple functions of HA has stimulated great interest in the involvement of HA in inflammatory diseases, tumorigenesis and metastasis.

Thus, there is a need to understand how HAS regulates HA size and how HA-UDP•HAS complexes can be stable for hours (DeAngelis and Weigel 1994; Kumari and Weigel 1997). Paradoxically, processive biosynthesis requires weak HA binding to HAS, which is likely different from the known stronger HA-binding proteins, to enable rapid on/off rates and HA-UDP chain movement within the enzyme. Each “new” HA-UDP must move from its binding site(s), yet not be free to diffuse or move away from the enzyme, so it can rebind to HAS and be added to new UDP-sugar(s) creating a new reducing end in the next catalytic cycle. We proposed (Tlapak-Simmons, Baggenstoss, Clyne, et al. 1999) that the processive mechanism of HAS utilizes an intra-HAS pore that traverses the plasma membrane and simultaneously traps (thus retaining) and translocates HA to the cell exterior in a continuous process as the chain is extended. An alternative ABC transporter model for HA transport does not explain why extracellular HA accumulation is not eliminated in streptococcal or mammalian cells, when such transporters are inhibited or knocked down (Ouskova et al. 2004; Prehm and Schumacher 2004). The ABC transporter mechanism for extracellular HA is refuted by a study concluding that it does not occur in human breast carcinoma cells (Thomas and Brown 2010) and by the recent demonstrations that purified S. equisimilis HAS (SeHAS) reconstituted in liposomes displays pore activity that mediates dye translocation (Medina et al. 2012) and also translocates HA to the lumen (Hubbard et al. 2012), as we originally proposed.

Several groups have made site-specific mutants of various Class I eukaryotic (Yoshida et al. 2000; Pummill and DeAngelis 2002, 2003) or streptococcal (Heldermon, Tlapak-Simmons, et al. 2001; Kumari et al. 2002, 2006; Kumari and Weigel 2005) HAS enzymes and shown effects on HA polymerizing activity or product size. In some mutants, HAS activity is decreased and in some cases HA product size can be decreased. Some recombinant HAS mutants produce smaller HA in the size ranges capable of cell signaling, supporting the possibility that similar changes in HAS activity or HA product size could be achieved in an allosteric-like manner in cells by either posttranslational modification or binding to regulator molecules that generate conformation states similar to that of a particular mutant.

Interestingly, the three different recombinant HAS isozymes produce HA of different sizes. COS-1 cell lines (Itano et al. 1999) expressing HAS1 and HAS3 secrete HA with size distributions from 0.2 to 2.0 MDa, whereas HAS2 cells make larger HA with an average molecular mass of >2 MDa. The three recombinant HAS isozymes synthesized >3.9 MDa HA in chinese hamster ovary cell lines (Brinck and Heldin 1999), but in isolated membranes HAS2 produced >3.9 MDa HA, whereas HAS3 made 0.1–1.0 MDa HA and HAS1 made HA of 0.12 MDa average mass. These studies as well as studies of streptococcal HASs indicate that the HA product size distribution made by any HAS is both intrinsic to the particular enzyme and also influenced by other factors, such as intracellular regulation, e.g. by phosphorylation (Goentzel et al. 2006; Bourguignon et al. 2007).

Despite the interest in how HAS regulates HA size, there have been no reports examining whether HA polymerizing activity and HA product size are tightly coordinated (i.e. coupled) or independent functions within the active site(s) of HAS. A further complication that makes it difficult to assess if activity and size control are independent functions is the use of different assay conditions to monitor HA synthesis activity and HA product size. Here, we investigated whether HAS activity and HA size are interdependent by investigating a panel of SeHAS mutants with different combinations of site-specific changes in four Cys residues conserved in the Class I family, particularly in mammals (Figure 1A). None of these four Cys residues are required for HAS activity, although their modification inhibits activity (Heldermon, Tlapak-Simmons, et al. 2001; Kumari et al. 2002; Kumari and Weigel 2005). Previous results showed that these four Cys residues are clustered together at the membrane interface and are within UDP-sugar substrate-binding sites (Figure 1B), indicating that they are well positioned to regulate either or both functions as HA is assembled and translocated to the cell surface. To avoid using two different assays to assess the two functions, we used SEC-MALLS (size exclusion chromatography–multi-angle laser light scattering) analysis to determine both weight-average HA mass and HAS activity simultaneously in the same samples. We found that HA product size is not coupled to or dependent on the specific polymerizing activity of HAS. The two enzyme functions can be altered independently.

Fig. 1.

Fig. 1.

Open in a new tab

A cluster of four conserved Cys residues in the active site(s) of Class I streptococcal and mammalian HAS proteins. (A) The four Cys residues in SeHAS, indicated by their position numbers, and the corresponding Cys residues in other streptococcal or mammalian enzymes are in boldface. HAS protein sequences are shown for: SeHAS (AAB87874); Streptococcus uberis (SuHAS, CAB46918); Streptococcus pyogenes HAS (SpHAS; AAA17981); human (HsHAS1, NP_001514; HsHAS2, NP_005319; HsHAS3, AF232772_1); mouse (MmHAS1, BAA11654; MmHAS2, AAC53309; MmHAS3, AAC53128); rat (RnHAS2, NP_037285); rabbit (OcHAS2, BAB63264; OcHAS3, BAB63265) and bovine (BtHAS2, CAA06239). (B) The model depicts the six membrane domains of SeHAS (numbered 1–6) and the close proximity of the four conserved Cys residues (circled) to each other and to the inner-membrane surface. The scheme is based on the SpHAS topology (Heldermon, DeAngelis, et al. 2001), and functional aspects of the Cys residues including their protection from modification by N-ethylmaleimide in the presence of UDP-sugars (Heldermon, Tlapak-Simmons, et al. 2001; Kumari et al. 2002; Kumari and Weigel 2005).

Results

Multiple factors can influence HA product size

In previous studies with a panel of SeHAS Cys mutants, we used a radioactive assay to quantify HA synthesis activity and agarose gel electrophoresis to monitor apparent HA product size (Kumari et al. 2002; Kumari and Weigel 2005). In order to assess a relationship between HA polymerizing activity and HA product size, we sought in the present study to eliminate differences in methods or conditions that might preferentially affect an assay for one of these characteristics. We and others have reported that many factors can affect the apparent size of HA products obtained with the same HAS preparation. We found, for example, that glycerol has different effects on different aspects of HAS stability and function. As with many enzymes, glycerol is a useful additive that enhances HAS activity and its storage lifetime (Itano et al. 1999; Tlapak-Simmons, Baggenstoss, Clyne, et al. 1999; Yoshida et al. 2000). However, HAS polymer synthesis is sensitive to the viscosity of its micro-environment, and the rate of sugar assembly is inhibited by increasing concentrations of glycerol or other viscogens (Tlapak-Simmons et al. 2004). We also find that when the glycerol content is increased from 2 to 20% (Figure 2), the HA products made by wild-type (WT) SeHAS are >60% smaller. Additionally, higher glycerol levels used to optimize activity are not tolerated well in MALLS analyses (Baggenstoss and Weigel 2006). Thus, the direct comparison of previous results for an effect on HAS activity and HA size was precluded by the use of different optimal glycerol concentrations in radioactive assays for activity vs in SEC-MALLS assays for HA size. Here, we eliminated problems associated with performing separate parallel assays by using SEC-MALLS, since each analysis provides, under the same experimental conditions in the same sample, the amount of HA present and the HA weight-average mass.

Fig. 2.

Fig. 2.

Open in a new tab

HA product size is decreased by increasing concentrations of glycerol. Membranes expressing WT or the indicated SeHAS mutant were incubated with substrates and 2% (black bars), 10% (gray bars) or 20% glycerol (white bars) at 30°C for 4 h, quenched and HA size was assessed by SEC-MALLS as described in Materials and methods.

HAS specific activity assessed using SEC-MALLS kinetic analysis

The utility and versatility of using SEC-MALLS data to assess HA product amount, weight-average mass, the size distribution (polydispersity) and the kinetics of synthesis is illustrated in Figure 3. For example, the refractive index (RI) profiles are different for the HA made by WT (black line) and C262/281A (red line) SeHAS (Figure 3A), with the mutant curve shifted (to the right) to smaller HA size. The calculated HA masses, in particular between 8 and 9 mL elution volume, show much smaller values for C262/281A (red circles) than for WT (black circles). Specific activities are determined by quantifying the amount of HA made (the area under the RI curve) by known amounts of the HAS protein. Differences in the amount of HA made between WT and four mutants (Figure 3B), using different quantities of membranes (so equal amounts of the HAS protein are present) reflect different specific activities of some mutant enzymes (e.g. WT and C262/281A make similar amounts of HA, whereas the C226S, C226/262A and Δ3C-281 mutants make less HA than WT). Figure 3C shows that during HA synthesis over 6 h, the areas of the RI profiles of products made by C226/262A (or WT or the other SeHAS mutants; not shown) increase and shift (to the left) to larger size products. This experiment also shows qualitatively that a steady-state HA size distribution is achieved by 4 h, even though the amount of HA continues to increase with time (e.g. at 5 and 6 h). Quantitative SEC-MALLS analysis confirmed that WT and the SeHAS mutants, including C226/262A and Δ3C-281 (Figure 4), reached a steady-state HA product size (weight-average molar mass, Mw) and size distribution by 4 h.

Fig. 3.

Fig. 3.

Open in a new tab

SEC-MALLS analysis for determining the kinetics of HA production and size changes and HAS specific activity. (A) Some HAS variants make HA of different mass than WT. The RI tracings for HA products made by WT SeHAS (black line) or the C262/281A mutant (red line) indicate the amount of HA present in the eluted volume. The calculated molar mass values in a subset of eluated volumes for the mutant (red circles) and WT (black circles) show a shift to smaller HA products made by the mutant. (B) Specific activity differences between WT and some mutants. The amount and masses of the HA made after 4 h of synthesis with equal amounts of the SeHAS protein (1 µg/mL reaction mix) are show by the RI profiles for WT (black), Δ3C-281 (green), C226S (blue), C226/262A (dark red) and C262/281A (red). Equal sample volumes (100 µL) were analyzed. (C) Simultaneous kinetic analysis of HA production and HA sizes made by SeHAS C226/262A. The changing RI traces reflect the size and amount of products HA in 100 µL reaction samples at 0.5 (dark green), 1.0 (light green), 2.0 (dark red), 3.0 (light red), 4.0 (gray), 5.0 (blue) and 6.0 (black) hours of synthesis. The series of chromatograms show (i) the increasing amounts of HA, (ii) the increasing (peak) size of the HA products and (iii) the broadening distribution of HA sizes over time.

Fig. 4.

Fig. 4.

Open in a new tab

Kinetics of HA Mw increase and attainment of steady-state HA product size distributions during HA biosynthesis. Samples were taken at 0.5, 1.0, 2.0, 3.0, 4.0, 5.0 or 6.0 h of HA synthesis by membranes containing SeHAS WT (A), C226/262A (B) or Δ3C-281 (C). The weight-average molar mass (Mw) of HA products was determined by SEC-MALLS analysis as described in Figure 3 and Materials and methods. Values are mean ± SD (n = 4–7).

The specific activities of SeHAS mutants were determined in two stages: (i) by SEC-MALLS kinetic analysis of increasing HA content (for replicate samples at multiple time points) from which rates of HA synthesis were calculated and (ii) by quantifying HAS protein content in membranes using a previously reported digital scanning method (Kumari et al. 2002) so that variable HAS expression levels were normalized in calculating the final specific HAS activity for each variant.

Effects of Cys mutagenesis on HA synthesis activity

The four Cys residues in SeHAS are absolutely conserved in the Class I streptococcal HASs and conserved positionally in mammalian Class I HASs (Figure 1A). Previous reports (Kumari and Weigel 1997, 2005; Heldermon, Tlapak-Simmons, et al. 2001; Kumari et al. 2002) showed that these Cys residues are not required for the function of SeHAS or Streptococcus pyogenes HAS, although their modification can decrease or ablate HAS activity. Two important findings (Kumari and Weigel 2005) were that the four Cys residues in SeHAS: (i) are clustered close together and (ii) are located at the membrane–HAS interface within the enzyme active site (Figure 1B). The involvement of these Cys residues in HAS activity and their high degree of conservation in Class I mammalian HASs promoted us to investigate their role in controlling the rate of HA synthesis and HA product size (Table I). Nine mutants showed no change in activity and nine mutants displayed significantly lower specific activities (from 36 to 85% of WT). Four mutants showed decreased specific activities that were <50% of WT; C226/262A (36%), Δ3C-281 (38%), C226A (44%) and C226S (45%). Mutants with altered rates of synthesis displayed normal linear kinetics of HA production (not shown).

Table I.

Effect of Cys mutagenesis on SeHAS-specific activity and HA product size

SeHAS Cys mutant Specific activity (percent of WT) HA Mw MDa (percent of WT)
WT 37.0 ± 4.0 (100) 3.78 ± 0.30 (100)
C226A 16.4 ± 3.9 (44)+ 3.29 ± 0.14 (87)**
C262A 29.1 ± 8.9 (79) 2.99 ± 0.29 (79)**
C281A 33.7 ± 6.0 (91) 2.71 ± 0.32 (72)***
C367A 46.7 ± 10.2 (126) 3.36 ± 0.43 (89)
C226S 16.5 ± 13.8 (45)** 2.84 ± 0.53 (75)*
C262S 31.4 ± 10.5 (85) 3.05 ± 0.46 (81)**
C281S 21.2 ± 2.0 (57)** 3.63 ± 0.43 (96)
C367S 29.5 ± 5.3 (80) 3.70 ± 0.55 (98)
C226/262A 13.4 ± 3.4 (36)*** 3.07 ± 0.27 (81)**
C226/281A 25.3 ± 4.3 (68)** 3.36 ± 0.44 (89)
C226/367A 37.8 ± 28.4 (102) 3.79 ± 0.80 (100)
C262/281A 30.2 ± 7.1 (82) 2.35 ± 0.26 (62)***
C281/367A 39.5 ± 4.9 (107) 3.40 ± 0.45 (90)
Δ3-C226 38.6 ± 9.6 (104) 2.60 ± 0.38 (69)*
Δ3-C262 31.6 ± 3.9 (85)* 3.56 ± 0.48 (94)
Δ3-C281 14.0 ± 1.9 (38)+ 3.44 ± 0.44 (91)
Δ3-C367 27.0 ± 3.3 (73)** 3.24 ± 0.63 (86)
C-Null 22.9 ± 9.8 (62)* 3.26 ± 0.44 (86)
Open in a new tab

Unlabeled HA was synthesized by membranes containing WT SeHAS or the indicated Cys mutants, and samples were analyzed for both the amount and the weight-average molar mass of HA made using SEC-MALLS as in experimental procedures. The size of HA made by SeHAS mutants varied relative to WT, but even the Cys-null mutant is active and makes near WT-sized HA. Three mutants (in bold) showed effects on both activity and size, and four mutants (in italics) showed no changes in either HA activity or size. The other 11 SeHAS mutants showed changes in only HA activity or HA size. Significant differences relative to WT are indicated by paired Student's t-test results: *P < 0.05; **P < 0.005; ***P < 0.0006; +P < 0.000003.

Effects of Cys mutagenesis on HA product size

More than half (10 of 18) of the Cys mutants showed no significant difference in HA size compared with WT and no mutants made larger HA (Table I). However, eight mutants made significantly smaller HA, ranging from 62 to 87% of WT size, as shown (Figure 3A) for the example of C262/281A. Three of the four single Cys-to-Ala mutants (C226A, C262A and C281A) and two of the four Cys-to-Ser mutants (C226S and C262S) made smaller HA. Two of the double Cys-to-Ala mutants tested, C226/262A and C262/281A, also made smaller HA. Only one of the four triple mutants (Δ3C-226) made smaller HA. SeHAS (C262/281A) made the smallest HA at 2.35 MDa. Interestingly, the Cys-null mutant made slightly smaller HA, but not significantly different than WT.

Functional association of Cys residues with mutant classification groups

Since individual mutants could show increased, decreased or no change in each of the two functions measured, they can be classified into nine possible groups based on the results in Table I. The 18 mutants fell into four of the nine possible groups (Table II); three groups were empty, since no mutants made larger HA. When grouped in this way, it was evident that each pattern of changes in HAS activity or HA size was associated with only, or predominantly, one or two of the four Cys residues. For example, C367 is the common altered Cys in Group 2 (C367S, C367A, C262/367A and C281/367A), whose members showed no changes in activity or size. Similarly, C226 and C281 are predominantly important for Group 3 (C281S, C226/281A, Δ3C-262, Δ3C-281, Δ3C-367 and Cys-null), in which activity is lower but HA size is unchanged. C262 and C281 are predominant in Group 5 (262A, 281A, 262S, C262/281A and Δ3C-226), whose members showed smaller HA but no change in HAS activity. The three mutants in Group 6, in which both polymerizing activity and HA size were decreased (C226S, C226A and C226/262A), have in common the alteration of C226.

Table II.

SeHAS Cys mutants fall into four of the nine categories

SeHAS Cys mutant group Effect on HA mass Effect on HAS specific activity No in group Mutant designations Cys residues associated with group
1 None Higher 0 — —
2 None None 4 367S; 367A; C226/367A; C281/367A C367
3 None Lower 6 281S; C226/281A; Δ3C-281; Δ3C-262; Δ3C-367; C-null C226 and C281
4 Smaller Higher 0 — —
5 Smaller None 5 262A; 281A; 262S; C262/281A; Δ3C-226 C262 and C281
6 Smaller Lower 3 226A; 226S; C226/262A C226
7 Larger Higher 0 — —
8 Larger None 0 — —
9 Larger Lower 0 — —
Open in a new tab

Since individual Cys mutants could show either larger, smaller or unaltered HA size and either higher, lower or unaltered HAS specific activity, each mutant can be classified into one of the nine possible groups (#1–9), based on the results in Table I. The 18 mutants examined fell into four of the nine possible groups. When organized in this way, it was evident that each pattern of changes in HAS activity and HA size is primarily associated with only one or two of the four Cys residues.

Discussion

It is generally assumed that UDP-sugar substrate availability can control HA product size; that smaller HA will be made if enzyme activity is low, due to low UDP-sugar substrate levels. Although this is a reasonable expectation based on the behavior of most enzymes, the Class I HASs are unique in being active as monomeric (Tlapak-Simmons et al. 1998; Pummill et al. 2001) phospholipid-dependent (Tlapak-Simmons, Baggenstoss, Clyne, et al. 1999) integral membrane proteins with multiple transmembrane domains (Heldermon, DeAngelis, et al. 2001), and in using UDP-sugars as acceptors rather than donors; the mechanism for sugar elongation at the reducing end is much different than that for typical glycosyltransferases (Weigel 2002). In fact, a [GlcNAc + GlcUA]-UDP disaccharide might be considered the reaction product of HAS, rather than a polymer of variable size. This distinction helps differentiate, as separate functions, the synthesis of HA disaccharide units from the control of HA product size. The present study tested this idea by assessing if these two aspects of HAS function can be uncoupled or if they always change in a linked or coordinated way.

Many known confounding factors can interfere with comparative assays of HAS activity and HA product size in vitro including the following. (i) HA size decreases as the amount of HAS increases (Baggenstoss and Weigel 2006). This is a mass action effect due to the altered substrate:enzyme ratio; even though more HA is made with more enzyme present, individual chains are smaller. (ii) Apparent reported size differences sometimes reflect the relatively long times required for HAS to complete one round of HA-UDP chain synthesis. Many studies do not consider that the synthesis of large chains (e.g. 4 MDa or ∼20,000 sugars) takes 2–4 h even under optimal conditions for linear synthesis (Baggenstoss and Weigel 2006). (iii) All Class I HASs are inhibited by released UDP, which then slows the rate of synthesis by competitive inhibition of UDP-sugar binding (Itano and Kimata 2002; Pummill and DeAngelis 2002; Tlapak-Simmons et al. 2004; Weigel and DeAngelis 2007). Due to the kinetic slowdown noted in (ii), this results in apparently smaller HA being made at later reaction times compared with identical samples containing alkaline phosphatase, which degrades UDP and helps maintain linearity (Baggenstoss and Weigel 2006). (iv) Most studies utilize whole cells or isolated cell membrane preparations in which HA synthesis is already in progress prior to starting an experiment (i.e. HA-UDP•HAS complexes are already present). In contrast, we express recombinant SeHAS in Escherichia coli SURE™ cells, which lack UDP-GlcUA, in order to start experiments with “naked” enzyme not associated with an HA-UDP chain.

If HA size control and HA elongation rate are inherently part of the same mechanism, then they should not be uncoupled by specific site-directed mutagenesis, unless the altered residue participates in both functions. The present results demonstrate that most of the SeHAS Cys mutants either displayed unaltered or decreased enzyme activity with no change in HA size or they made smaller HA with no change in activity. Thus, uncoordinated changes in one of the two functions occurred in 11 of the 18 mutants examined, whereas only 3 mutants showed coordinated changes, in which both activity and size decreased, and 4 mutants showed no changes in either activity or size. The results demonstrate that different combinations of the four conserved Cys residues participate in each of these two important HAS functions—sugar assembly rate and regulation of size. Cys367 can be altered with no consequence to either (Group 1, Table II), whereas the mutation of only Cys226 affects both activity and size (Group 4, Table II). Interestingly, changing Cys226 in combination with Csy281 restores normal HA size (Group 3, Table II).

The intraprotein pore explains the ability of HAS to stay bound to HA-UDP in a processive manner for hours, but how does HAS regulate HA product size? We propose a retain–release model (Figure 5) to explain how glycosyl polymerases such as HAS make large products and why a polymer product is not released until a certain characteristic size range is attained (Heldermon et al. 2000). In this model, the forces acting on HA are represented by vectors (arrows). Multiple diverse interactions between HA-UDP and HAS, within and surrounding the intraprotein pore, provide individual binding forces (e.g. H-bonds and ionic and hydrophobic interactions) and a strong topological constraint that contribute to an overall (net) retention force (Fretain), whose direction points through the pore to the cell interior. Fretain serves to keep the HA-UDP•HAS complex together (i.e. processive), despite an opposing overall net release force (Frelease) that acts in the opposite direction (i.e. through the pore pointing to the exterior) to pull the HA-UDP chain away (i.e. dissociation or release). Frelease results from multiple types of forces in the external environment that act on HA, including random Brownian motion, fluid currents, cell surface or extracellular matrix proteins that bind HA (increasing mass and perhaps anchoring the HA), and inertial forces due to cell or surface membrane movements.

Fig. 5.

Fig. 5.

Open in a new tab

The HA•HAS retain–release model for the control of HA product size by Class I HAS. The left scheme shows the six membrane domains of SeHAS organized within the cell membrane (large dark rectangle) creating a pore (Medina et al. 2012) that encircles and binds to the growing HA chain containing alternating GlcUA (red circles) and GlcNAc (blue circles) attached to UDP (yellow triangle) at the reducing end. The dashed arrows show the two substrates, UDP-GlcUA and UDP-GlcNAc, feeding into the intracellular active sites as the UDP reaction products leave. The small arrows (light green) denote various forces that contribute to the binding between HA and HAS or associated phospholipids (not shown) and create an overall net retention force (Fretain) that keeps the HA-UDP•HAS complex together (large light green arrow pointing down). The small dark green arrows represent interactions (e.g. binding to external proteins or matrix) and forces (e.g. shear generated by rotational or translation cell movement or Brownian motion) that act to pull the HA chain free. The large green arrow (pointing up) indicates the overall net force acting to release the HA from HAS (Frelease). Fretain remains relatively constant during HA biosynthesis, whereas Frelease will naturally increase as HA size increases. Initially (left scheme) Fretain >> Frelease and as synthesis proceeds, they eventually become equal (middle of scheme), until finally at some point during chain elongation (right scheme), an HA size will be attained for which Frelease > Fretain and the HA chain will be pulled from the HAS pore and released to the medium. The released chain is shown as HA-UDP, although it is not known if the physical release process is associated with UDP loss and the release of free HA. Two interacting MD residues (K48 in MD2 and E327 in MD4) that contribute to Fretain are indicated in the right scheme in yellow letters (Kumari et al. 2006).

Based on the present results showing that the synthesis rate and the HA size are not coupled, Fretain will be essentially constant during the synthesis of a growing HA chain; the local forces and catalytic activities within an HA-UDP•HAS complex would not change as HA size increases. In contrast, Frelease will increase as HA length increases, thus increasing the probability of HA chain release as HA mass increases. Frelease will be random and continuously fluctuate, and it will be substantial only when the vector components of many forces transiently align along the pore axis to create a large sum for Frelease. Since 1.1 MDa HA (∼5850 sugars) has a particle diameter of ∼200 nm and HAS is likely <10 nm, it is easy to visualize that as HA gets bigger, it can be pulled away from HAS more easily.

A distribution of HA sizes around an optimum mass is inherently created in this model because HA release is a stochastic event. The exact time and particular HA mass during synthesis when a transient Frelease will be great enough to release the HA will vary for each HA-UDP•HAS complex and the external micro and macro conditions. Thus, different size HA products will be released, even from the same HAS protein as it makes successive HA chains under the same conditions. The retain–release model predicts that HA size will increase if Fretain within HA-UDP•HAS increases or if Frelease decreases. Minimizing or decreasing flow forces acting on a large HA chain should allow HAS to produce larger HA, whereas more severe forces should facilitate earlier release and give smaller HA (i.e. by shortening biosynthesis time, not by cleaving the HA chain).

The model also predicts that HAS mutations, posttranslational modifications or conformation changes, induced by other factors, that weaken and decrease Fretain within HA-UDP•HAS will cause the enzyme to release smaller HA. For example, the inherent HA product size differences among HAS isozymes noted above can be explained by intrinsically different retention forces generated within the three HA-UDP•HAS complexes because their primary structures are ∼30% different. Our earlier finding that mutations of K48 in MD2 and E327 in MD4 of SeHAS can decrease HA size by >80% (Kumari et al. 2006) is consistent with the presence of these residues within the HAS pore and indicates that they contribute substantially to the overall Fretain. In contrast and consistent with their location near the “inner mouth” of the pore, the current results indicate that the four conserved cysteines contribute less to the overall Fretain, since HA size is decreased in only a small subset of mutants and the decreases are not as great (<38%).

The pore model for HA synthesis, involving a HAS-lipid complex (Tlapak-Simmons, Baggenstoss, Clyne, et al. 1999, 2005; Weigel 2004; Kumari et al. 2006; Medina et al. 2012), is consistent with the location of the important cluster of four conserved Cys residues at the membrane interface and within substrate-binding sites (Figure 1B). The active sites are likely positioned near the membrane–enzyme junction so that as polymer elongation occurs, the chain is threaded into and subsequently pushed/pulled through the intra-HAS pore, simultaneously providing a mechanism for both processive HA synthesis and HA chain translocation. Additional mechanistic and structural studies are needed to test the pore and retain–release models at the atomic level and to understand the role of Cys and other residues in the processes by which HAS assembles disaccharide units and independently translocates and releases its HA product across the membrane to the cell exterior.

Materials and methods

Materials and Cys mutants

Media components were from Difco (New Thermo Fisher Scientific, Boston, MA). Streptomyces hyaluronicus hyaluronidase, UDP-GlcUA, UDP-GlcNAc, calf intestinal alkaline phosphatase and other reagent grade chemicals were from Sigma-Aldrich (St Louis, MO). The generation of single, double, triple and null Cys mutants of SeHAS was described previously (Kumari et al. 2002). Escherichia coli SURE 2 Cell cultures were grown and membranes were prepared as reported previously (Tlapak-Simmons, Baggenstoss, Kumari, et al. 1999).

HAS assays

Membrane pellets were suspended by brief sonication at 4°C and added to 2-mL microcentrifuge tubes to a final concentration of 10–20 μg/mL total membrane protein in assay buffer containing 18.75 mM sodium phosphate and 6.25 mM potassium phosphate, pH 7.0, 75 mM NaCl, 2% (v/v) glycerol, 0.1 mM ethylene diamine tetra-acetic acid (EDTA), 1.0 mM dithiothreitol, 0.1 mM phenylmethylsulfonyl fluoride, 1.0 μM pepstatin and 2 μM leupeptin. UDP-GlcUA and UDP-GlcNAc were added to 1 mM final concentrations, and alkaline phosphatase was added (Baggenstoss and Weigel 2006) to a final concentration of 0.02 U/μL. The synthesis reactions were then initiated by the addition of MgCl2 to a final concentration of 20 mM and incubated for 4 h, unless stated otherwise, at 30°C in a vibrating Taitec mixer (San Jose, CA). Synthesis was terminated by adding EDTA to a final concentration of 40 mM, chilling on ice for 20 min and then heating to 65–75°C in a 100°C sand bath (1 min per 0.1 mL volume) to inactivate HAS and phosphatase. Samples were extracted with three volumes of CHCl3:MeOH (2:1) by shaking at 22°C for 15 min, centrifuged at 6800 × g for 5 min and the aqueous layers were removed and residual organic solvent was removed by Speed Vacuum Centrifugation (Savant, Holbrook, NY). The samples were then analyzed by SEC-MALLS, which is a static (rather than a dynamic or quasi-elastic) light scattering system.

Determination of HA weight-average mass

SEC-MALLS analyses were performed by chromatographic separation of samples (200 µL) using one PLaquagel-OH60 column (Polymer Labs; Agilent Tech., Santa Rosa, CA) alone or in series with a PLaquagel-OH40 column with or without a second PLaquagel-OH60 (depending on the HA size range of the samples) at 22°C at a flow rate of 0.4–0.5 mL/min in 50 mM sodium phosphate, pH 7.0, 150 mM NaCl, 0.05% sodium azide. MALLS analysis was performed continuously on the column eluate as it passed through a DAWN DSP Laser Photometer in series with an OPTILAB DSP Interferometric Refractometer (both from Wyatt Technologies, Inc., Santa Barbara, CA). Data were analyzed using Astra v4.73, a dn/dc value of 0.153, an A2 value of 0.0023 and either first-order Zimm or second-order Berry fits, depending on the size of the HA being analyzed. In our experience, first-order Zimm fits have less uncertainty in analysis of data for HA of <2 MDa, whereas second-order Berry analysis of data for HA of >2 MDa gives better fits. Column performance was monitored by the routine analysis of a commercial HA (2.4 MDa; Genzyme, Cambridge, MA) used as a reference. Based on our experience with SEC-MALLS, we avoid using HA concentrations >50 µg/mL, since artificially high mass values may occur, possibly due to chain entanglement. MALLS measures RI that inherently provides concentration data, so samples can be easily monitored in order to stay below this concentration and minimize concentration-dependent artifacts. Controls with EV membranes confirmed that no MALLS signals were present in the regions where HA was detected, unless SeHAS was present.

Determination of HAS specific activity

The SEC-MALLS RI values collected for the above samples were integrated to obtain the total amount of HA present in each sample. Replicate samples were analyzed at different times of synthesis to verify linearity (with time and membrane protein content) and to determine the rate of increase in HA products (i.e. HA amount made per hour). In order to normalize for the different levels of SeHAS expression among the different Cys mutants, E. coli membranes containing WT or mutant SeHAS were solubilized and subjected to sodium dodecylsulfate-polyacrylamide gel electrophoresis in 10% (w/v) gels (Kumari et al. 2002). The SeHAS protein in each membrane preparation was quantified by the image analysis of Coomassie stained gels using a Fluorchem™ 8000 (Alpha Innotech Corp., San Leandro, CA). Integrated values for SeHAS bands in membranes were compared with values for total protein in each lane to estimate the SeHAS protein content per milligram of membrane protein. These data were then used to normalize the amount of HA produced to the amount of the HAS protein in the assay (µg HA/µg SeHAS/h).

Funding

This research was supported by National Institute of General Medical Sciences grant GM35978 from the National Institutes of Health.

Conflict of interest

None declared.

Abbreviations

EDTA, ethylene diamine tetra-acetic acid; GlcNAc, N-acetyl-d-glucosamine; GlcUA, d-glucuronic acid; HA, hyaluronic acid/hyaluronan; HAS, HA synthase; RI, refractive index; SEC-MALLS, size exclusion chromatography–multi-angle laser light scattering; SeHAS, Streptococcus equisimilis HAS; WT, wild type.

Acknowledgements

We thank Amy Padgett-McCue and Jennifer Washburn for technical assistance.

References

  1. Baggenstoss BA, Weigel PH. Size exclusion chromatography-multiangle laser light scattering analysis of hyaluronan size distributions made by membrane-bound hyaluronan synthase. Anal Biochem. 2006;352:243–251. doi: 10.1016/j.ab.2006.01.019. doi:10.1016/j.ab.2006.01.019. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Bourguignon LY, Gilad E, Peyrollier K. Heregulin-mediated ErbB2-ERK signaling activates hyaluronan synthases leading to CD44-dependent ovarian tumor cell growth and migration. J Biol Chem. 2007;282:19426–19441. doi: 10.1074/jbc.M610054200. doi:10.1074/jbc.M610054200. [DOI] [PubMed] [Google Scholar]
  3. Brinck J, Heldin P. Expression of recombinant hyaluronan synthase (HAS) isoforms in CHO cells reduces cell migration and cell surface CD44. Exp Cell Res. 1999;252:342–351. doi: 10.1006/excr.1999.4645. doi:10.1006/excr.1999.4645. [DOI] [PubMed] [Google Scholar]
  4. DeAngelis PL, Weigel PH. Immunochemical confirmation of the primary structure of streptococcal hyaluronan synthase and synthesis of high molecular weight product by the recombinant enzyme. Biochemistry. 1994;33:9033–9039. doi: 10.1021/bi00197a001. doi:10.1021/bi00197a001. [DOI] [PubMed] [Google Scholar]
  5. Goentzel BJ, Weigel PH, Steinberg RA. Recombinant hyaluronan synthase 3 is phosphorylated in mammalian cells. Biochem J. 2006;396:347–354. doi: 10.1042/BJ20051782. doi:10.1042/BJ20051782. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Heldermon C, DeAngelis PL, Weigel PH. Topological organization of the hyaluronan synthase from Streptococcus pyogenes. J Biol Chem. 2001;276:2037–2046. doi: 10.1074/jbc.M002276200. doi:10.1074/jbc.M002276200. [DOI] [PubMed] [Google Scholar]
  7. Heldermon C, Kumari K, Tlapak-Simmons VL, Weigel PH. Streptococcal hyaluronan synthases and the synthesis of designer hyaluronan. In: Abatangelo G, Weigel PH, editors. New Frontiers in Medical Sciences: Redefining Hyaluronan. Amsterdam: Elsevier Science, B.V; 2000. pp. 41–50. [Google Scholar]
  8. Heldermon CD, Tlapak-Simmons VL, Baggenstoss BA, Weigel PH. Site-directed mutation of conserved cysteine residues does not inactivate the Streptococcus pyogenes hyaluronan synthase. Glycobiology. 2001;11:1017–1024. doi: 10.1093/glycob/11.12.1017. doi:10.1093/glycob/11.12.1017. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Hubbard C, McNamara JT, Azumaya C, Patel MS, Zimmer J. The hyaluronan synthase catalyzes the synthesis and membrane translocation of hyaluronan. J Mol Biol. 2012;418:21–31. doi: 10.1016/j.jmb.2012.01.053. doi:10.1016/j.jmb.2012.01.053. [DOI] [PubMed] [Google Scholar]
  10. Itano N, Kimata K. Mammalian hyaluronan synthases. IUBMB Life. 2002;54:195–199. doi: 10.1080/15216540214929. doi:10.1080/15216540214929. [DOI] [PubMed] [Google Scholar]
  11. Itano N, Sawai T, Yoshida M, Lenas P, Yamada Y, Imagawa M, Shinomura T, Hamaguchi M, Yoshida Y, Ohnuki Y, et al. Three isoforms of mammalian hyaluronan synthases have distinct enzymatic properties. J Biol Chem. 1999;274:25085–25092. doi: 10.1074/jbc.274.35.25085. doi:10.1074/jbc.274.35.25085. [DOI] [PubMed] [Google Scholar]
  12. Kumari K, Baggenstoss BA, Parker AL, Weigel PH. Mutation of two intramembrane polar residues conserved within the hyaluronan synthase family alters hyaluronan product size. J Biol Chem. 2006;281:11755–11760. doi: 10.1074/jbc.M600727200. doi:10.1074/jbc.M600727200. [DOI] [PubMed] [Google Scholar]
  13. Kumari K, Tlapak-Simmons VL, Baggenstoss BA, Weigel PH. The streptococcal hyaluronan synthases are inhibited by sulfhydryl modifying reagents but conserved cysteine residues are not essential for enzyme function. J Biol Chem. 2002;277:13943–13951. doi: 10.1074/jbc.M110638200. doi:10.1074/jbc.M110638200. [DOI] [PubMed] [Google Scholar]
  14. Kumari K, Weigel PH. Molecular cloning, expression, and characterization of the authentic hyaluronan synthase from group C Streptococcus equisimilis. J Biol Chem. 1997;272:32539–32546. doi: 10.1074/jbc.272.51.32539. doi:10.1074/jbc.272.51.32539. [DOI] [PubMed] [Google Scholar]
  15. Kumari K, Weigel PH. Identification of a membrane-localized cysteine cluster near the substrate binding sites of the Streptococcus equisimilis hyaluronan synthase. Glycobiology. 2005;15:529–539. doi: 10.1093/glycob/cwi030. doi:10.1093/glycob/cwi030. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Medina AP, Lin J, Weigel PH. Hyaluronan synthase mediates dye translocation across liposomal membranes. BMC Biochem. 2012;13:2. doi: 10.1186/1471-2091-13-2. doi:10.1186/1471-2091-13-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Noble PW, McKee CM, Cowman M, Shin HS. Hyaluronan fragments activate an NF-kappa B/I-kappa B alpha autoregulatory loop in murine macrophages. J Exp Med. 1996;183:2373–2378. doi: 10.1084/jem.183.5.2373. doi:10.1084/jem.183.5.2373. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Ouskova G, Spellerberg B, Prehm P. Hyaluronan release from Streptococcus pyogenes: Export by an ABC transporter. Glycobiology. 2004;14:931–938. doi: 10.1093/glycob/cwh115. doi:10.1093/glycob/cwh115. [DOI] [PubMed] [Google Scholar]
  19. Prehm P, Schumacher U. Inhibition of hyaluronan export from human fibroblasts by inhibitors of multidrug resistance transporters. Biochem Pharmacol. 2004;68:1401–1410. doi: 10.1016/j.bcp.2004.06.017. doi:10.1016/j.bcp.2004.06.017. [DOI] [PubMed] [Google Scholar]
  20. Pummill PE, DeAngelis PL. Evaluation of critical structural elements of UDP-sugar substrates and certain cysteine residues of a vertebrate hyaluronan synthase. J Biol Chem. 2002;277:21610–21616. doi: 10.1074/jbc.M202456200. doi:10.1074/jbc.M202456200. [DOI] [PubMed] [Google Scholar]
  21. Pummill PE, DeAngelis PL. Alteration of polysaccharide size distribution of a vertebrate hyaluronan synthase by mutation. J Biol Chem. 2003;278:19808–19814. doi: 10.1074/jbc.M301097200. doi:10.1074/jbc.M301097200. [DOI] [PubMed] [Google Scholar]
  22. Pummill PE, Kempner ES, DeAngelis PL. Functional molecular mass of a vertebrate hyaluronan synthase as determined by radiation inactivation analysis. J Biol Chem. 2001;276:39832. doi: 10.1074/jbc.M105489200. doi:10.1074/jbc.M105489200. [DOI] [PubMed] [Google Scholar]
  23. Stern R, Asari AA, Sugahara KN. Hyaluronan fragments: An information-rich system. Eur J Cell Biol. 2006;85:699–715. doi: 10.1016/j.ejcb.2006.05.009. doi:10.1016/j.ejcb.2006.05.009. [DOI] [PubMed] [Google Scholar]
  24. Thomas NK, Brown TJ. ABC transporters do not contribute to extracellular translocation of hyaluronan in human breast cancer in vitro. Exp Cell Res. 2010;316:1241–1253. doi: 10.1016/j.yexcr.2010.01.004. [DOI] [PubMed] [Google Scholar]
  25. Tlapak-Simmons VL, Baggenstoss BA, Clyne T, Weigel PH. Purification and lipid dependence of the recombinant hyaluronan synthases from Streptococcus pyogenes and Streptococcus equisimilis. J Biol Chem. 1999;274:4239–4245. doi: 10.1074/jbc.274.7.4239. doi:10.1074/jbc.274.7.4239. [DOI] [PubMed] [Google Scholar]
  26. Tlapak-Simmons VL, Baggenstoss BA, Kumari K, Heldermon C, Weigel PH. Kinetic characterization of the recombinant hyaluronan synthases from Streptococcus pyogenes and Streptococcus equisimilis. J Biol Chem. 1999;274:4246–4253. doi: 10.1074/jbc.274.7.4246. doi:10.1074/jbc.274.7.4246. [DOI] [PubMed] [Google Scholar]
  27. Tlapak-Simmons VL, Baron CA, Gotschall R, Haque D, Canfield WM, Weigel PH. Hyaluronan biosynthesis by Class I streptococcal hyaluronan synthases occurs at the reducing end. J Biol Chem. 2005;280:13012–13018. doi: 10.1074/jbc.M409788200. doi:10.1074/jbc.M409788200. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Tlapak-Simmons VL, Baron CA, Weigel PH. Characterization of the purified hyaluronan synthase from Streptococcus equisimilis. Biochemistry. 2004;43:9234–9242. doi: 10.1021/bi049468v. doi:10.1021/bi049468v. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Tlapak-Simmons VL, Kempner ES, Baggenstoss BA, Weigel PH. The active streptococcal hyaluronan synthases (HASs) contain a single HAS monomer and multiple cardiolipin molecules. J Biol Chem. 1998;273:26100–26109. doi: 10.1074/jbc.273.40.26100. doi:10.1074/jbc.273.40.26100. [DOI] [PubMed] [Google Scholar]
  30. Weigel PH. Functional characteristics and catalytic mechanisms of the bacterial hyaluronan synthases. Int Union Biochem Mol Biol. 2002;54:201–210. doi: 10.1080/15216540214931. doi:10.1080/15216540214931. [DOI] [PubMed] [Google Scholar]
  31. Weigel PH. The hyaluronan synthases. In: Garg H, Hale CA, editors. Chemistry and Biology of Hyaluronan. Oxford, UK: Elsevier, Ltd; 2004. pp. 553–567. [Google Scholar]
  32. Weigel PH, DeAngelis PL. Hyaluronan synthases: A decade-plus of novel glycosyltransferases. J Biol Chem. 2007;282:36777–36781. doi: 10.1074/jbc.R700036200. doi:10.1074/jbc.R700036200. [DOI] [PubMed] [Google Scholar]
  33. West DC, Hampson IN, Arnold F, Kumar S. Angiogenesis induced by degradation products of hyaluronic acid. Science. 1985;228:1324–1326. doi: 10.1126/science.2408340. doi:10.1126/science.2408340. [DOI] [PubMed] [Google Scholar]
  34. Yoshida M, Itano N, Yamada Y, Kimata K. In vitro synthesis of hyaluronan by a single protein derived from mouse HAS1 gene and characterization of amino acid residues essential for the activity. J Biol Chem. 2000;275:497–506. doi: 10.1074/jbc.275.1.497. doi:10.1074/jbc.275.1.497. [DOI] [PubMed] [Google Scholar]

Articles from Glycobiology are provided here courtesy of Oxford University Press

RESOURCES