Structural Requirements for the Stability and Microsomal Transport Activity of the Human Glucose 6-Phosphate Transporter*

Deficiencies in glucose 6-phosphate (G6P) transporter (G6PT), a 10-helical endoplasmic reticulum transmembrane protein of 429 amino acids, cause glycogen storage disease type 1b. To date, only three missense mutations in G6PT have been shown to abolish microsomal G6P transport activity. Here, we report the results of structure-function studies on human G6PT and demonstrate that 15 missense mutations and a codon deletion (ΔF93) mutation abolish microsomal G6P uptake activity and that two splicing mutations cause exon skipping. While most missense mutants support the synthesis of G6PT protein similar to that of the wild-type transporter, immunoblot analysis shows that G20D, ΔF93, and I278N mutations, located in helix 1, 2, and 6, respectively, destabilize the G6PT. Further, we demonstrate that G6PT mutants lacking an intact helix 10 are misfolded and undergo degradation within cells. Moreover, amino acids 415–417 in the cytoplasmic tail of the carboxyl-domain, extending from helix 10, also play a critical role in the correct folding of the transporter. However, the last 12 amino acids of the cytoplasmic tail play no essential role(s) in functional integrity of the G6PT. Our results, for the first time, elucidate the structural requirements for the stability and transport activity of the G6PT protein.

tions resulting from infections due to chronic neutropenia and functional deficiencies of neutrophils and monocytes (6,7). Thus, GSD-1b offers a potential model for exploring the molecular mechanism(s) of neutropenia, neutrophil/monocyte dysfunction, and its relationship to G6Pase deficiency.
Complementary DNAs encoding human (8), mouse (9), and rat (9) G6PT have been isolated and characterized. Two adjacent lysine residues at the C termini of mammalian G6PT proteins conform to an ER membrane protein retention motif, consistent with the cellular location of the transporter. Recently, the orientation of human G6PT in the ER has been resolved by protease protection and glycosylation scanning analyses (10). Those studies showed that the G6PT protein is anchored to the ER by 10 transmembrane helices with both N and C termini facing the cytoplasm.
Human G6PT, a single copy gene located on chromosome 11q23 (11), spans approximately 5.5 kilobases and is composed of 9 exons (12,13). Using G6Pase-deficient mice, we have shown that the transport and hydrolysis of G6P are tightly coupled processes and that G6Pase activity is required for efficient G6P transport into the microsomes (14). Based on that finding, a functional assay for the recombinant G6PT protein was established (12), which showed that G6PT can function minimally as a G6P transporter in the absence of G6Pase. However, microsomal G6P uptake activity was markedly enhanced in the simultaneous presence of G6PT and G6Pase (12).
To date, a large number of mutations have been identified in the G6PT gene of GSD-1b patients (8,12,(15)(16)(17)(18)(19)(20)(21)(22). However, only three disease-causing missense mutations have been functionally characterized (12). The resulting three amino acid substitutions, R28H, G149E, and C183R, abolish microsomal G6P transport activity as shown by heterologous expression of mutant proteins in COS-1 cells. Here we report the identification of four novel G6PT mutations among 14 GSD-1b patients. Fifteen missense mutations and a codon deletion mutation completely abolish microsomal G6P transport activity, and two splicing mutations cause exon skipping. Our studies establish that mutations in the G6PT gene cause GSD-1b.
At present, very little is known about the structural requirements for the correct folding and G6P transport activity of the G6PT protein. In this study, we undertake structure-function studies and demonstrate that G20D, ⌬F93, and I278N mutations, located in helix 1, 2, and 6, respectively, destabilize the transporter. Further, we show that intactness of helix 10 in G6PT is essential for proper folding and stability of the transporter. However, the last 12 amino acids in the cytoplasmic tail of the C-domain in G6PT, including the ER transmembrane protein retention motif, are not essential for membrane insertion, stability, or microsomal G6P transport activity. Finally, we establish the minimal length of the cytoplasmic tail of the C-domain essential for functional integrity of the G6PT.

MATERIALS AND METHODS
Mutational Analysis-We have analyzed the G6PT gene of 13 GSD-1b patients that presented with hypoglycemia, hepatomegaly, growth retardation, neutropenia, neutrophil dysfunction, and recurrent bacterial infections; each also lacks a genetic defect in the G6Pase gene. We have also analyzed the G6PT gene of one GSD-1 patient diagnosed clinically as GSD-1a who proved to have no mutation in the G6Pase gene. The lymphoblast line of GSD-1b patient 1 reported by Beaudet et al. (6) was obtained from the National Institute of General Medical Sciences, Human Genetic Mutant Cell Repository (Camden, NJ). Genomic DNA preparations were extracted from lymphoblasts or blood samples using a Nucleon II kit (Scotlab Bioscience). All peripheral blood samples were obtained with the informed consent of the patients and/or parents.
The G6PT gene in GSD-1 patients was characterized by single strand conformation polymorphism analysis (23) on MDE (mutation-detectionenhancement nondenaturing) gels (AT Biochem, Malvern, PA) containing 5% glycerol. Exon-containing fragments were amplified by polymerase chain reaction using primers containing intronic, 5Ј-untranslated, and 3Ј-untranslated sequences of the human G6PT gene as described previously (12). The mutation-containing fragments identified by single strand conformation polymorphism were subcloned and characterized by DNA sequencing.
Construction of G6PT Mutants and Expression in COS-1 Cells-The template for G6PT mutant construction by PCR was nucleotides 166 -1496 of the human G6PT cDNA in a pSVL vector (Amersham Pharmacia Biotech), which contains the entire coding region at nucleotides 170 -1459, with the translation initiation codon, ATG, at nucleotides 170 -172 (8). Human G6PT contains a BstEII site at nucleotides 863-869 (8). The PCR primers for mutants that contained mutations located upstream of the BstEII site are nucleotides 166 -186 (sense) and 851-878 (antisense) in the G6PT cDNA. The primers for mutants that contained mutations located downstream of the BstEII site are nucleotides 851-878 (sense) and 1476 -1496 (antisense). The amplified fragments were ligated into either the pSVLhG6PT-BstEII-3Ј or the pSVLhG6PT-BstEII-5Ј fragment.
The mutant primers are as follows: G20D (nucleotides 218 - The G6PT-5ЈFLAG and G6PT-3ЈFLAG constructs (10) containing the 8-amino acid FLAG marker peptide, DYKDDDDK (Scientific Imaging Systems, Eastman Kodak Co.) at the 5Ј-and 3Ј-end of the G6PT coding region, respectively, were used as templates for constructing the FLAGtagged mutants. The nucleotide sequence in all constructs was verified by DNA sequencing.
COS-1 cells were grown at 37°C in HEPES-buffered Dulbecco's modified minimal essential medium supplemented with 4% fetal bovine serum. The G6PT construct in a pSVL vector was transfected into COS-1 cells by the DEAE-dextran/chloroquine method (24) in the absence or presence of a co-transfected G6Pase cDNA as described previously (12). After incubation at 37°C for 2 days, the transfected cultures were harvested for microsomal G6P uptake assay, Western blot analysis, or RNA isolation.
Northern Blot, Western Blot, and in Vitro Transcription-Translation Analyses-Total RNA was isolated by the guanidinium thiocyanate/ CsCl method (25), fractionated by electrophoresis through a 1.2% agarose gel containing 2.2 M formaldehyde, and transferred to a Nytran membrane by electroblotting. The membranes were hybridized with either a uniformly labeled G6PT or ␤-actin riboprobe.
For Western blot analysis of FLAG-tagged G6PT, microsomal proteins were separated by electrophoresis through a 12% polyacrylamide-SDS gel and trans-blotted onto polyvinylidene fluoride membranes (Millipore Corp., Bedford, MA). The membranes were incubated with a monoclonal antibody against the FLAG epitope (Scientific Imaging Systems). The immunocomplex was detected with the horseradish peroxidase-linked chemiluminescent system containing the SuperSignal West Pico Chemiluminescent substrate obtained from Pierce.
In vitro transcription-translation of G6PT cDNA constructs in a pGEM-7Zf(ϩ) vector was performed using the TnT coupled reticulocyte lysate system obtained from Promega Biotech (Madison, WI). L-[ 35 S]methionine was used as the labeled precursor. The in vitro synthesized proteins were analyzed by 12% polyacrylamide-SDS gel electrophoresis and fluoroautoradiography.
G6P Uptake Assays-G6P uptake measurements were performed essentially as described previously (12). Briefly, microsomes (40 g) were incubated in a reaction mixture (100 l) containing 50 mM sodium cacodylate buffer, pH 6.5, 250 mM sucrose, and 0.2 mM [U-14 C]G6P (50 Ci/mol). The reaction was stopped at the appropriate time by filtering immediately through a nitrocellulose membrane (BA85, Schleicher & Schuell) and washed with an ice-cold solution containing 50 mM Tris-HCl, pH 7.4 and 250 mM sucrose. Microsomes permeabilized with 0.2% deoxycholate, which abolished G6P uptake, were used as negative controls. Two to three independent experiments were conducted, and at least three G6P uptake studies were performed for each microsomal preparation. We have previously shown that microsomal G6P uptake in COS-1 cells transfected with a G6PT and a G6Pase cDNA reached a plateau at 3 min of incubation (12). Therefore, the 3-min G6P uptake values were used to compare microsomal G6P uptake activities of wild-type (WT) and mutant G6PT proteins. Statistical analysis using the unpaired t test was performed with The GraphPad Prism Program (GraphPad Software, San Diego, CA). Data are presented as the mean Ϯ S.E.

Mutations Identified in the G6PT Gene of GSD-1b Patients-
Single strand conformation polymorphism and DNA sequencing analyses were used to identify mutations in the G6PT gene of 13 GSD-1b patients and one putative GSD-1a patient, although the G6Pase gene had no identifiable mutation in the coding region. G6PT mutations were identified in all 14 patients, indicating that the GSD-1a patient was misdiagnosed. Fourteen different mutations were identified, including nine missense (G20D, R28C, L85P, W118R, G149E, G150R, I278N, G339C, and A393D), two nonsense (W96X and W393X), one splicing (550ϩ2t 3 g), and two deletion (1211delCT and ⌬F93) mutations (Table I). Four novel mutations identified in this study are L85P, ⌬F93, I278N, and A373D.
G6PT Mutations That Cause GSD-1b-To date, 59 different mutations, including 23 missense, have been identified in the G6PT gene of GSD-1b patients (Refs. 8, 12, and 15-22 and this study). However, only three missense mutations, R28H, G149E, and C183R, uncovered in this laboratory were shown to abolish microsomal G6P transport activity (12). To demonstrate that other missense mutations uncovered in GSD-1b patients cause the disease, we constructed a series of G6PT mutants and examined the ability of each mutant to transport G6P in transient expression studies. We also examined G6P transport activity of the ⌬F93 and N198I mutants; the latter was identified as a possible polymorphic marker in a GSD-1b patient homozygous for a splicing mutation (15).
The combined transfection of the G6Pase and WT G6PT cDNAs into COS-1 cells shows that G6P was efficiently taken up by intact microsomes isolated from these cells (Table II). In contrast, microsomal G6P transport activity was undetectable when G20D, R28C, S55R, G68R, ⌬F93, L85P, G88D, W118R, G150R, I278N, R300H, G339C, or A393D mutant G6PT cDNA was co-transfected with the G6Pase cDNA (Table II). The N198I construct retained at least 95% of WT G6P transport activity (Table II), indicating that it represents a sequence polymorphism in the G6PT gene.
The G20D, ⌬F93, and I278N Mutations Destabilize the Transporter-Northern blot analysis confirmed that similar levels of G6PT transcripts were expressed in WT or mutant G6PT-transfected COS-1 cells (Fig. 1A). Our data, therefore, demonstrate that the decrease in G6P uptake was due to a defective G6PT protein and not due to a decrease in efficiency of expression of the transfected cDNA construct.
We also constructed G6PT mutants containing the FLAG tag at the C terminus and examined their biosynthesis after tran-sient expression in COS-1 cells (Fig. 3A). Again G20D-3ЈFLAG, ⌬F93-3ЈFLAG, and I278N-3ЈFALG constructs supported the synthesis of markedly reduced amounts of G6PT proteins as compared with the WT-3ЈFLAG as well as the R28H-3ЈFLAG   a Microsomal G6P uptake activity from WT or mutant G6PT cDNA-transfected cells in the presence of a co-transfected G6Pase cDNA was measured after a 3-min incubation and is presented as the mean Ϯ S.E. Statistical analysis was performed using the unpaired t test.

constructs.
In vitro transcription-translation assays showed that WT, R28H, G2OD, ⌬F93, and I278N constructs directed the synthesis of similar amounts of G6PT proteins (Fig. 3B). Therefore, the observed decrease in the synthesis of G20D, ⌬F93, and I278N mutants may result from misfolding and rapid degradation of mutant proteins in a heterologous expression system.
The 550ϩ2t3g and 550ϩ1g3t Mutations Cause Exon Skipping-The 550ϩ2t3g splicing mutation (Table I) is likely to cause exon-2 skipping from the G6PT transcript. Sequencing of six G6PT cDNA clones obtained by reverse transcriptase-PCR of RNA isolated from lymphoblasts of patient 7 showed that exon 2 (233-bp) sequences were absent from half of the cDNA clones. The other half contained the 228G3 A/G20D mutation.
We have previously identified a splicing mutation (550 ϩ 1g3t) in a compound heterozygous patient also harboring a 716T3 C/C183R mutation (12). To demonstrate that the 550ϩ 1g3t mutation also causes exon 2 skipping from the G6PT transcript, we sequenced five G6PT cDNA clones obtained by reverse transcriptase-PCR of RNA isolated from lymphoblasts of the patient. While three of the cDNA clones contained the 716T3 C/C183R mutation, the other two clones lacked exon 2 sequences, demonstrating that the 550ϩ1g3t mutation also causes exon-2 skipping.

Roles of the C-terminal Cytoplasmic Tail and Helix 10 in Folding and Stability of the G6PT-
The cytoplasmic tail at amino acids 415-429 in the C-domain of human G6PT is extended directly from helix 10, which encompasses amino acids 396 -414 (Fig. 2). Two nonsense mutations, W393X and R415X (19), lacking helix 10 as well as the cytoplasmic tail and the entire cytoplasmic tail, respectively, were identified in the G6PT gene of GSD-1b patients. To investigate the molecular basis of G6PT deficiency caused by these mutations, we examined G6PT synthesis directed by W393X and R415X constructs (Fig. 4A). While the W393X-5ЈFLAG construct supported little or no G6PT synthesis following transfection into COS-1 cells, the R415X-5ЈFLAG construct supported the synthesis of a significant amount of G6PT, approximating 30 -40% of the levels in WT-5ЈFLAG-transfected cells. This suggests that helix 10 is required for the correct folding and stability of the transporter. We further examined G6PT synthesis directed by two additional mutants, E401X-5ЈFLAG and T408X-5ЈFLAG, that disrupt helix 10. Again, we observed little or no G6PT synthesis following transfection of these mutants into COS-1 cells (Fig. 4A). Northern blot analysis showed that similar levels of G6PT transcripts were expressed in WT-, W393X-, E401X-, T408X-, and R415X-transfected cells (Fig. 4B). Moreover, the four mutant constructs, like the WT G6PT, directed the synthesis of similar amounts of G6PT proteins in a cell-free transcriptiontranslation system (Fig. 4C). Our data indicate that helix 10 in G6PT also plays a vital role for correct folding of the transporter and that the incorrectly folded mutants were degraded in the cell.
G6PT contains an ER transmembrane protein retention signal, KKAE, at the C terminus (Fig. 2). The observed reduction in the synthesis of the R415X mutant in COS-1 cells prompted us to investigate the role of this signal on folding and stability of G6PT. Interestingly, K420X-5ЈFLAG, K426X-5ЈFLAG, and K427X-5ЈFLAG mutants lacking this signal supported the synthesis of similar amounts of G6PT protein as the WT construct (Fig. 4A). This was supported by Northern blot analysis showing that similar levels of G6PT transcripts were expressed in WT as well as mutant constructs transfected with COS-1 cells (Fig. 4B). Again, WT as well as mutant constructs directed the synthesis of similar amounts of G6PT proteins in a cell-free transcription-translation system (Fig. 4C).
The Length of the Cytoplasmic Tail Essential for Folding and Stability of the Transporter-Our results show that the R415X mutant lacking the entire 15-amino acid cytoplasmic tail in the C-domain was degraded more rapidly than the WT construct in the COS-1 heterologous expression system. On the other hand, the K420X construct lacking the last 10 amino acids in the cytoplasmic tail had similar stability as the WT G6PT. To determine the minimal length of cytoplasmic tail required for G6PT folding and/or stability, we examined G6PT synthesis in transient expression studies directed by 5ЈFLAG-tagged G6PT mutants that lack the last 14 (N416X), 13 (I417X), 12 (R418X), and 11 (T419X) amino acids in the tail (Fig. 5A). While N416X and I417X supported the synthesis of reduced levels of G6PT proteins compared with the WT construct, R418X, T419X, like K420X, supported the synthesis of similar amounts of proteins as the WT G6PT (Fig. 5A). The decrease in G6PT synthesis by R415X, N416X, and I417X constructs appeared to correlate to the length of the cytoplasmic tail retained by these mutants.
Northern blot analysis showed that similar levels of G6PT transcripts were expressed in WT as well as in N416X-, I417X-, R418X-, and T419X-transfected COS-1 cells (Fig. 5B). Likewise, all directed the synthesis of similar amounts of G6PT proteins in a cell-free transcription-translation system (Fig. 5C).
G6P Uptake Activity of G6PT Mutants-The lack of detectable amounts of mutant G6PT in W393X-transfected COS-1 cells could explain functional G6PT deficiencies manifested by GSD-1b patients harboring this mutation. On the other hand, a significant amount of mutant protein was found in R415Xtransfected cells. We therefore examined the ability of the various G6PT mutants to transport G6P in transient expression assays in the presence of a co-transfected G6Pase cDNA (Table III). W393X, E401X, and T408X mutant constructs were devoid of microsomal G6P uptake activity, reflecting the lack of G6PT protein in mutant-transfected cells. On the other hand, microsomal G6P transport activity in R415X-, N416X-, and I417X-transfected cells were approximately 47, 78, and 80% of the activity in G6PT-WT-transfected cells (Table III), indicating that these three G6PT mutants are fully functional in microsomal G6P uptake. As expected, microsomal G6P transport activities of R418X, T419X, K420X, K426X, and K427X mutants were indistinguishable from that of the WT construct (Table III).

DISCUSSION
In this report, we have undertaken structure-function studies of the G6PT, and our results show that G2OD, ⌬F93, and I278N mutations, located in helix 1, 2, and 6, respectively, destabilize the transporter. Further, we demonstrate that helix 10, encompassing amino acids 396 -414 in G6PT and amino acids 415-417 in the cytoplasmic tail of the C-domain extending directly from helix 10, also play important roles in folding and stability of the transporter. On the other hand, the last 12 amino acids of the cytoplasmic tail, including the ER transmembrane retention signal, are not required for the correct folding or microsomal G6P transport function. Additionally, we have characterized the G6PT gene in 14 GSD-1b patients and uncovered four novel mutations. To demonstrate that G6PT mutations identified in GSD-1b patients cause the disease, we generated a series of mutant G6PT constructs carrying missense and the ⌬F93 mutation and examined microsomal G6P transport activity in transient expression assays. In earlier studies, we have shown that G6P transport in intact microsomes from cells transfected with the R28H, G149E, or C183R construct was undetectable (12). We now show that G20D, R28C, S55R, G68R, L85P, G88D, ⌬F93, W118R, G150R, I278N, R300H, G339C, and A393D mutants were devoid of microsomal G6P transport activity. Our data firmly establish that GSD-1b is caused by mutations in the G6PT gene that abolish microsomal G6P transport function of the encoded protein.
Among the 16 codon mutations that we have characterized to date, G20D, ⌬F93, and I278N mutant constructs support the synthesis of markedly reduced amounts of G6PT proteins when compared with the WT construct in a heterologous expression system. However, WT as well as G20D, ⌬F93, and I278N transcripts direct the synthesis of similar amounts of G6PT proteins in a cell-free translation system. Since folding of nascent proteins occurs during their in vitro synthesis on rabbit reticulocyte ribosomes, our data strongly suggest that substitution of glycine 20 with an aspartic acid in helix 1, deletion of phenylalanine 93 in helix 2, or substitution of isoleucine 278 with an asparagine in helix 6 ( Fig. 2) causes misfolding of the mutant protein. The observed decrease in the synthesis of G20D, ⌬F93, and I278N mutant proteins in COS-1 cells was caused by the misfolded proteins being targeted by the degradative apparatus in the cell before being translocated to the ER membrane. Degradation of abnormal or incompletely synthesized proteins represents a quality control system in the cell (26). It has been shown that the major cystic fibrosis mutation, ⌬F508, leads to defective protein folding and subsequent degradation of the misfolded mutant (27). The total absence of G6P transport activity in microsomes of G20D-, ⌬F93-, and I278N-transfected COS-1 cells indicates that the low levels of these mutant proteins are nonfunctional.
It has been demonstrated that co-translational folding begins from the N terminus and proceeds to the C terminus (28,29). For many proteins, the C-terminal segment of 20 -30 amino acid residues, which is sheltered by the ribosome prior to the release of the full-length polypeptide, is essential for formation of the native, biologically active structure. In this study, we show that T408X, E401X, and W393X mutants that lack the cytoplasmic tail in the C-domain as well as a disrupted (T408X and E401X) or a truncated (W393X) helix 10 were degraded when expressed in COS-1 cells. Yet WT as well as mutant transcripts directed the synthesis of similar amounts of G6PT proteins in a cell-free translation system. This indicates that helix 10, encompassing amino acids 396 -414, and the cytoplasmic tail of G6PT are important for the folding of the transporter into an active configuration. The lack of G6PT protein synthesis directed by these three mutant constructs in a heterologous expression system may have resulted from the failure to fold completely and subsequent degradation of the incompletely folded proteins.
We show that truncations of amino acids 415-417 in the cytoplasmic tail of G6PT that extends directly from helix 10 lead to a decrease in the amounts of mutant proteins synthesized in COS-1 cells and a corresponding reduction in microsomal G6P transport activity. The similar amounts of mutant and WT proteins synthesized in a cell-free system indicate that amino acids 415-417 may also play a role in the correct folding and/or stability of the G6PT. On the other hand, truncations of any of the last 12 amino acids in the cytoplasmic tail, including the ER transmembrane protein retention signal, had no effect on synthesis and microsomal G6P transport activity of the transporter, indicating that these residues are not essential for functional integrity of the transporter. We have previously shown that G6Pase, another ER transmembrane protein, remains enzymatically active and associated with the microsomes following truncations of its C-terminal 8 amino acids, including the ER retention signal (30). It appears that the intactness of their transmembrane helices is vital for membrane insertion and functions of G6PT as well as G6Pase.
The R415X mutation was identified in a heterozygous state in two GSD-1b patients of Afro-Caribbean origin (19). No additional mutation was found in any of the exons, intron 1, and the 500-bp promoter region of the G6PT gene. The authors speculated that the putative second mutation, if it exists, must greatly reduce the expression of that copy of the G6PT gene (19). Although it is possible that the R415X mutant protein is less stable in vivo than our in vitro expression system, the synthesis of significant amounts of a functional mutant G6PT following transfecting the R415X construct into COS-1 cells strongly argues that the G6PT genes of these two patients warrant further investigation.
In summary, we have elucidated a number of the structural requirements for the stability and microsomal transport function of the transporter and demonstrated that 15 missense mutations, the ⌬F93 deletion mutation, and two splicing G6PT mutations cause GSD-1b. a Microsomal G6P uptake activity from WT or mutant G6PT cDNA-transfected cells in the presence of a co-transfected G6Pase cDNA was measured after a 3-min incubation and is presented as the mean Ϯ S.E. Statistical analysis was performed using the unpaired t test.