The Characteristics of Liver Glucose-6-phosphatase in the Envelope of Isolated Nuclei and Microsomes Are Identical*

We have compared the characteristics of glucose-6- phosphatase (EC 3.1.3.9) in the envelope of purified nuclei and microsomes from rat liver. The latency of mannose-6-P hydrolysis, permeability to EDTA, and susceptibility of the enzyme to protease-mediated in- activation all indicated that the permeability barrier defined by the envelope in situ is significantly disrupted in isolated nuclei (i.e. in vitro). Latency of mannose-6-P hydrolysis was demonstrated to provide a quantitative measure of the degree of nuclear mem- brane disruption. Electron micrographs confirmed the existence of substantial regions of the envelope in vitro where the permeability barrier to EDTA was intact (Le. an “intact component”). The kinetics of glucose-6-phosphatase catalyzed by the intact component was obtained by subtracting the contribution of enzyme in disrupted regions from the total enzymic activity of untreated nuclei. The characteristics of glucose-6- phosphatase in intact and fully disrupted membranes of nuclei were indistinguishable from microsomes with respect to (a) the kinetics of glucose-6-P hydrolysis, (b) the effects of incubations with mannose-6-P, N-ethylmaleimide, and protease from Bacillus amyloli- quefaciens, (c) the extremely high latency of carbamyl phosphate:glucose phosphotransferase activity, and (d) both the patterns of response of activity and the change

We have compared the characteristics of glucose-6phosphatase (EC 3.1.3.9) in the envelope of purified nuclei and microsomes from rat liver. The latency of mannose-6-P hydrolysis, permeability to EDTA, and susceptibility of the enzyme to protease-mediated inactivation all indicated that the permeability barrier defined by the envelope in situ is significantly disrupted in isolated nuclei (i.e. in vitro). Latency of mannose-6-P hydrolysis was demonstrated to provide a quantitative measure of the degree of nuclear membrane disruption. Electron micrographs confirmed the existence of substantial regions of the envelope in vitro where the permeability barrier to EDTA was intact (Le. an "intact component"). The kinetics of glucose-6phosphatase catalyzed by the intact component was obtained by subtracting the contribution of enzyme in disrupted regions from the total enzymic activity of untreated nuclei. The characteristics of glucose-6phosphatase in intact and fully disrupted membranes of nuclei were indistinguishable from microsomes with respect to (a) the kinetics of glucose-6-P hydrolysis, (b) the effects of incubations with mannose-6-P, Nethylmaleimide, and protease from Bacillus amyloliquefaciens, (c) the extremely high latency of carbamyl phosphate:glucose phosphotransferase activity, and (d) both the patterns of response of activity and the change in latency of glucose-6-phosphatase induced by fasting, experimental diabetes, and cortisol injection.
Our results show clearly that apparent differences in the glucose-6-phosphatase activity of untreated preparations of nuclei and microsomes are simply expressions of significant differences in the degree of intactness of their respective permeability barriers. Since flattened cisternae, characteristic of the rough endoplasmic reticulum in situ, are preserved in intact regions of the envelope of isolated nuclei, the present findings constitute the most direct and definitive evidence to date that the properties of glucose-6-phosphatase in the endoplasmic reticulum in situ are faithfully reproduced with intact microsomes.
Conflicting perceptions prevail as to the role of the ER' * This work was supported in part by Research Grant AM-19625 from the National Institutes of Health. A preliminary report describing part of these studies has appeared (1). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
$ To whom correspondence should be addressed.
3 Predoctorial trainee supported by National Institutes of Health 'The abbreviations used are: ER, endoplasmic reticulum; TI, a glucose-6-P-specific translocase that mediates the penetration of the membrane in the function of glucose-6-phosphatase (EC 3.1.3.9). A key point of contention concerns the mechanisms by which the membrane influences the degree of latency' of the associated phosphatase and phosphotransferase activity. Arion and coworkers (2)(3)(4)(5) have demonstrated that glucose-6-P hydrolysis by intact microsomes3 involves the coupled functions of three integral components of the ER membrane: 1) a glucose-6-P-specific translocase, denoted TI, that mediates penetration of the hexose phosphate into the luminal cavity, 2) a relatively nonspecific phosphohydrolase with its active site at the luminal surface, and 3) a second translocase, denoted TP, that mediates P, efflux as well as slow rates of penetration of PPI and carbamyl-P through the membrane.
Analyses from several laboratories indicate that glucose-6-P hydrolysis is the principal and probably the exclusive function of glucose-6-phosphatase of intact organelles in uiuo (4, 10-12). The specificity of the enzyme in uiuo with respect to its substrate and function appears to be restricted to glucose-6-P hydrolysis by a host of factors which include: (a) the selective transport properties of TI and T2 (2,4), (b) the physiologic concentrations of glucose-6-P, Pi, glucose, PPi, and carbamyl-P (4),4 and (c) the intrinsic kinetic characteristics of the coupled system which limit its capacity to catalyze net glucose-6-P synthesis (4, 6). 4 Latency is the percentage of activity in fully disrupted microsomes that is not expressed in untreated or intact microsomes. Latency is calculated as 100 (activity in disrupted membranesactivity in intact or untreated membranes)/activity in disrupted membranes. In this paper, microsomes or nuclei isolated from liver homogenates, washed and assayed without further treatment are referred to as "untreated." As noted previously (4)(5)(6)8), untreated microsomes are heterogeneous preparations composed of intact vesicles ("intact microsomes") in which the limiting membrane acts as a selective selective permeability barrier and disrupted structures in which selective permeability is lacking and the enzyme has free access to ionic substrates and inhibitors. The proportion of the two forms is easily quantified by assays of the "low K," mannose-6-phosphatase activity that is expressed only in disrupted structures (5,7,8). The enzymic activity of intact microsomes is calculated as the activity of untreated microsomes minus the contribution of enzyme in the disrupted component (4)(5)(6). Untreated microsomes are converted to fully disrupted of the physiologic function of glucose-6-phosphatase which advocates an important role for the associated phosphotransferase activity (see Refs. 13-16 for reviews). These investigators have proposed that the relatively high latency of the phosphotransferase activity observed with untreated3 homogenates and microsomes arises through imposition of artificial constraints on the enzyme or translocases as a consequence of the morphological changes that occur when microsomes are generated from fragments of the ER. In support of this thesis they have reported that the phosphohydrolase and phosphotransferase activities are significantly more latent in microsomes than in isolated hepatic nuclei (17)(18)(19) or in rat hepatocytes rendered permeable to ionic substrates by exposure to filipin (20,21). However, the relevance of their observations rests on the validity of the underlying assumption that the native structure of the ER and, therefore, the function of glucose-6-phosphatase in situ are retained in the envelope of isolated nuclei and in filipin-treated hepatocytes.
The "integrity" or degree of "intactness" of the selective permeability barrier defined by the single limiting membrane of hepatic microsomes can be quantified either by measuring EDTA permeability (8, 22) or more conveniently by determining the latency of the "low K, " mannose-6-P phosphohydrolase activity catalyzed by glucose-6-phosphatase (5, 7, 8). These criteria are now used routinely in studies with hepatic and renal microsomes. By contrast, except for the apparent preservation of gross morphology seen in electron micrographs of isolated nuclei (18), the intactness of the permeability barrier defined by the nuclear envelope has not been evaluated. Likewise, no information is available as to either the structure or permeability characteristics of the ER in hepatocytes exposed to filipin. Thus, it remained to be established whether studies with either of these experimental models of the "ER in situ" generate reliable data on the characteristics of glucose-6-phosphatase in the intact cell. The characteristics of the nuclear glucose-6-phosphatase system are presented here. Results of a parallel study of filipintreated rat hepatocytes will be presented elsewhere.s EXPERIMENTAL PROCEDURES~ RESULTS A N D DISCUSSION Quantitative Assessment of the "Intactness" of the Envetope of Isolated Nuclei-Mannose-6-P phosphohydrolase activity routinely was between 40 and 60% latent in our nuclear preparations (Table I and below). In our experience maximal expression of nuclear mannose-6-phosphatase activity and other activities catalyzed by glucose-6-phosphatase required exposure of nuclear preparations to 0.1% sodium deoxycholate rather than the 0.05% concentration routinely used by others in their characterizations of the enzyme in nuclear membranes (17-19). The activity of mannose-6-phosphatase in nuclei supplemented with 0.05% deoxycholate averaged only 72.0 f 5.8% ( n = 32; range = 58.9 to 82.7%) of the value determined with preparations supplemented to 0.1% deoxy-

TABLE I
Correlation between the latency of mannose-6-phosphatase and EDTA impermeability in the envelope of isolated rat liver nuclei Mannose-6-phosphatase was assayed at pH 6.5 with 1 mM substrate. EDTA permeability was quantified by assessing its ability to solubilize Pi from lead phosphate precipitates formed during prior incubation of nuclei with 1 mM glucose-6-P and 2 mM Pb(NO& (see under "Experimental Procedures" in the miniprint). cholate. Although only data obtained with the higher concentration of detergent are presented herein, the influences of both concentrations on nuclear activities were examined throughout the present study.
Cytochemical studies revealed a high degree of correlation between the latency of mannose-6-phosphatase and the percentage of lead phosphate precipitated in the perinuclear space that was inaccessible to EDTA (Table I). The electron micrograph in Fig. 1 shows that after exposure to EDTA the lead phosphate deposits were confined to the aspect of the nucleus defined by the envelope and not in membrane vesicles that may have contaminated our nuclear preparations, e.g. microsomes or vesicles derived from the nuclear membranes (18). The micrograph confirms the existence in the nuclear envelope of substantial regions of membrane which were impermeable to EDTA.
Further validation of the relationship between these indices of membrane integrity is presented in the experiment detailed in Fig. 2 and the accompanying text. 6 A Model of the Envelope of Isolated Nuclei-The characteristics of mannose-6-phosphatase and EDTA permeability strongly suggest that the envelope of isolated nuclei does not define a continuous permeability barrier. We believe that the preceding results and those that follow are most readily explained in terms of the schematic drawn in Fig. 3. Ultrastructural studies of intact cells show clearly that the membranes of the ER and nuclear envelope constitute a morphologic continuum in situ and that the continuum is disrupted when tissues are homogenized (23-28). Whether the envelope in an isolated nucleus defines an intact or a disrupted permeability barrier is likely determined by the location of the sites at which disruption occurs between the ER proper and the outer aspect of the envelope. If the location is in a region of the ER proper (e.g. a t points designated by arrows labeled A in the upper panel of Fig. 3), the broken membranes associated with the envelope would have the freedom to contact one another and fuse. Disruption and fusion in this case would generate a region of the envelope bounded by a continuous permeability barrier, as illustrated by the intact area in the lower panel of Fig. 3, and it would account for the "blebs" that are occasionally seen in the envelope of isolated nuclei (23, 27). Alternatively, if disruption occurs a t loci on the envelope proper (e.g. at arrows labeled B in the upper panel of Fig. 3), it is expected that the physical restriction imposed by association of the inner aspect of the envelope (the so-called inner membrane) with the chromatin network (23, 24, 26) would prevent association and, therefore, fusion of the broken membranes. This would generate a region in the envelope where the permeability barrier is interrupted (designated leaky in the lower panel of Fig. 3). The components localized within this region, which is the perinuclear cisterna in situ, would have free but artificial access to ionic metabolites (e.g. mannose-6-P and EDTA) and macromolecules (see below). Such interrupted regions (i.e. gaps) in the envelope have frequently been noted in electron micrographs of isolated nuclei (23, 25-28).
The scheme in Fig. 3 formed the basis for our experimental approach to the characterization of glucose-6-phosphatase in isolated nuclei. Implicit in the model is the existence of two "glucose-6-phosphatase activities" in untreated nuclei, one activity representing the coupled system (the intact component) and the other the enzyme in disrupted membranes. Accordingly, the kinetics of glucose-6-P hydrolysis in microsomes and nuclei was compared in fully disrupted and intact membranes. The latter data were obtained by subtracting the contribution of the "disrupted component" from the activities determined for untreated membrane preparations as detailed under "Experimental Procedures" in the Miniprint Competitive Interactions of Nuclear Glucose-6-phosphatase with Mannose-6-P and Glucose-6-P-The experiment summarized in Table I1 was carried out to provide an additional test of the proposition that latency of mannose-6-P hydrolysis provides a reliable quantitative measure of the degree of intactness of the nuclear envelope. The competitive interactions between mannose-6-P and glucose-6-P were studied before and after supplementing nuclei to 0.1% deoxycholate, i.e. in untreated and fully disrupted membranes. The data for fully disrupted membranes define the situation in which both substrates have complete and equal access to the active sites B. Mannose-6-P hydrolysis of the phosphohydrolase. Whereas glucose-6-P was equally effective in inhibiting mannose-6-P hydrolysis in untreated and disrupted preparations, mannose-6-P was substantially less effective in inhibiting glucose-6-P hydrolysis by untreated compared with fully disrupted nuclei. If it is assumed that mannose-6-P interacted only with the enzyme in disrupted membranes (i.e. 48% of the total enzyme), the observed inhibition of glucose-6-P hydrolysis catalyzed by the disrupted component in untreated nuclei (45%) agrees well with the expected value (39%) calculated from the 82% inhibition of glucose-6-phosphatase in fully disrupted nuclei.
The observations with rat liver nuclei reproduce earlier Glucose-6-phosphatase in the Nuclear Envelope findings with microsomes from rat liver (7) and human liver ( 5 ) . They strongly support the validity of using latency of mannose-6-P hydrolysis to quantify the intactness of the nuclear envelope in hepatocytes.
Influences of Protease Treatment on Glucose-6-phosphate Activity in Microsomes and Nuclei-Dallner and coworkers (3,29,30) have used proteolytic enzymes to probe and define the transverse topology of the ER membrane. The finding that glucose-6-phosphatase can be inactivated only when the protease can penetrate into the luminal compartment (e.g. after treatment with deoxycholate) demonstrates that the active site of the enzyme is situated at the luminal surface. Measurements of EDTA permeability and latency of mannose-6-P hydrolysis (Table I) suggested that approximately 50% of the glucose-6-phosphatase in the nuclear envelope is localized in "leaky" membranes and accordingly should be accessible to added proteases. Therefore, the influences of protease on glucose-6-phosphatase of isolated nuclei were studied with two objectives in mind 1) to verify the susceptibility of enzyme in Ieaky membranes to proteolytic degradation and 2) to compare the responses of the glucose-6phosphatase system in intact microsomes and the intact nuclear envelope to proteolytic treatments.
The procedure for treatment with protease was first evaluated in studies with intact microsomes and microsomes rendered freely permeable to macromolecules by prior exposure to 0.1 M NH,OH (3) (see Table 111 and accompanying text). The effects of protease treatment on nuclear glucose-6-phosphatase are summarized in Table IV. The increased latency of mannose-6-P hydrolysis following exposure to the protease was expected. It is predictable from Fig. 3 that protease treatment would preferentially destroy the enzyme in leaky membranes and thus increase the proportion of enzyme localized within intact regions of the recovered membranes. The effects of protease treatment on the kinetics of glucose-6-P

TABLE IV
The effect of treatment with protease on nuclear glucose-6phosphatase activity Assay media, pH 6.5, contained in 1 ml: 50 mM Tris/cacodylate buffer, 10 mg of bovine plasma albumin (defatted), nuclei before and after supplementation with 0.1% deoxycholate, and the indicated concentrations of phosphate substrates. Other details are given under ''Experimental Procedures" in the Miniprint. The mean values from 2 independent experiments are presented. Because of the obvious influence of protease treatment on milligrams of nuclear protein, activities are expressed as "recovered milliunits" rather than units/mg of protein (see Table 111).
Activity values in intact nuclear membranes were normalized t o correspond to that expected if all the enzyme in fully disrupted membranes was housed in intact membranes (also see under "Experimental Procedures" in the Miniprint).
2 m M glucose-6-P Estimates of K,, mM hydrolysis were qualitatively and quantitatively similar to those observed with microsomes (Table 111); protease treatment increased the apparent K, for glucose-6-P in intact but not disrupted membranes. This was accompanied by increased latency of glucose-6-phosphatase activity, especially in media containing 2 m M glucose-6-P.
Comparison of the Effect of N-Ethylmaleimide on Microsomal and Nuclear Glucose-6-phosphatase Activity-Exposure of microsomes to NEM inhibits glucose-6-P hydrolysis by intact microsomes (31, 32). The inhibition is abolished when NEM-treated microsomes are subsequently disrupted (e.g. with detergents). At neutral pH, 1 mM or less of the sulfhydryl poison causes no inhibition when incubated with the enzyme of fully disrupted microsomes (31, 32).7 Inhibition by NEM is expressed as an increase in the Michaelis constant for glucose-6-P without a change in V,,, (31). Moreover, the inhibition can be specifically prevented by the presence of glucose-6-P at concentrations greater than 50 m~~; mannose-6-P, Pi, and D-glucose have no protective influence at concentrations as high as 0.2 M. These observations indicate that NEM reacts with a thiol group located at or near the glucose-6-P-binding site on the translocase. Table V compares the influence of NEM on glucose-6phosphatase of nuclei and microsomes. The thiol poison had no effect on glucose-6-phosphatase activity in preparations that were disrupted after exposure to NEM, and the latency of mannose-6-phosphatase was unaltered by NEM exposure.
When activity values obtained in the absence of deoxycholate supplementation were corrected for the contribution of disrupted component, the latency of glucose-6-P hydrolysis was identical in intact microsomes and intact nuclear membranes before NEM treatment, and the thiol poison caused identical inhibitions in nuclei and microsomes.
Carbamyl-P:Glucose Phosphotransferase Activity-Gunderson and Nordlie (17, 18) have concluded that all phosphohydrolase and phosphotransferase activities catalyzed by glucose-6-phosphatase are much less "constrained (i.e. less latent) in nuclei than in microsomes. The data in Table VI compare the activity levels of carbamyl-P:glucose phosphotransferase in nuclei and microsomes before and after disruption with optimal concentrations of sodium deoxycholate. The results confirm that carbamyl-Pglucose phosphotransferase, like mannose-6-phosphatase, appears much less latent in untreated nuclei as compared with untreated microsomes. However, when the phosphotransferase activities in the untreated membranes are adjusted for the contribution of enzyme in leaky membranes, virtually identical and extremely high latencies are observed for the two preparations.
Some Nutritional and Endocrine Effects on Glucose-6-phosphatase in Rat Liver Nuclei-The influences of changes in nutritional and endocrine status on glucose-6-phosphatase from isolated rat liver nuclei were studied in response to earlier conclusions that (a) there are fundamental differences in influence of these variables on the enzyme from nuclei and microsomes (19) and (b) that 15 to 20% of the total hepatic glucose-6-phosphatase is localized in the nuclear envelope (18, 19). Numbers of animals in each group, body and liver weights, blood glucose concentrations, total liver and nuclear DNA, and protein values are presented in Table VII. Treatment regimens are described under "Experimental Procedures" in the Miniprint. Assays were performed on both whole liver homogenates and nuclei isolated from the homogenates. Homogenates were supplemented with deoxycholate to 0.1, 0.2, and 0.3%, although only the data obtained with the Latency and sensitivity to N-ethylmaleimide of glucose-6-phosphatase in nuclear and microsomal membranes before and after correcting for the disrupted component The means of three determinations in each of two experiments are presented. See under "Experimental Procedures" in "miniprint" for details of NEM treatment. Assay conditions were as described in Table 11, except 1 mM glucose-6-P was used.

TABLE VI
Comparison of carbamyl-P:glucose phosphotransferase activities in hepatic nuclei and microsomes from a 24-h fasted rat Assay media, pH 7.0, contained in a volume of 1.0 ml: 50 mM Tris/cacodylate buffer, 10 mg of bovine plasma albumin, 100 mM glucose, and 5 mM carbamyl-P (phosphotransferase assays only) or 1 mM mannose-6-P (phosphatase assays), and microsomes before and after supplementation to 0.2% deoxycholate (144 and 36 pg of protein, respectively) or nuclei before and after supplementation to 0.1% deoxycholate (205 and 103 pg of protein, respectively). Microsomes and nuclei were isolated from the same liver homogenate. The means of 3 determinations are presented. optimal concentration of detergent (ie. 0.2%) are given. Results are presented for phosphohydrolase activities observed before correcting for the contribution of the disrupted component (Table VIII) and after such corrections were made (Table IX).

Membrane preparation Latency
Mannose-6-P hydrolysis was at least 94% latent in whole homogenates, and its latency was not significantly altered by the changes in endocrine or nutritional status of the rats (Table VIII, part B). In confirmation of earlier findings (reviewed in Refs. 13 and 33), the latency of glucose-6-phosphatase activity was markedly influenced by changes in endocrine and nutritional status. Latency of glucose-6-phosphatase in liver homogenates was increased by fasting and especially diabetes, whereas cortisol injection caused only a modest and statistically insignificant increase in Iatent activity.
In all groups, the phosphohydrolase activities in isolated nuclei were only about one-half as latent as observed for the homogenates (Table VIII, part B). However, when activity values for glucose-6-P hydrolysis by untreated preparations were corrected for the disrupted component, essentially identical responses to treatment are seen in intact membranes of nuclei and homogenates (Tabel IX).
The percentage of nuclei recovered from whole homogenates was determined by analyses of DNA based on the assumption that 98.5% of the liver DNA is localized in the nucleus (34). Mean recoveries ranged from 39% for 24-h fasted rats to 67% for diabetic animals (Table VII). These recoveries are similar to those reported by others (26,27,35). When the percentage of recovery of nuclear DNA was used to normalize the nuclear phosphohydrolase activities to 100% recovery, the data (Table IX, part C) show that only about 2% of the total hepatic glucose-6-phosphatase is localized in the nuclear envelope. The much higher value reported previously (18,19) appears to have originated from a calculation using an estimate (from Ref. 36) of the protein content of a crude nuclear fraction (ie. a 600 x g pellet) rather than purified nuclei. Our findings support the conclusion of Kartenbeck et al. (26) that the nuclear envelope can be quantitatively neglected in consideration of the overall rate of glucose formation from glucose-6-P in the rat hepatocyte.
The smallest percentage of total liver activity was seen in nuclei from the streptozotocin-diabetic rats. Garfield and Cardell (37) have reported that experimental diabetes increases the proportion of total hepatic glucose-6-phosphatase localized in the smooth ER. Therefore, the lower fraction of liver glucose-6-phosphatase in the nuclei from diabetic rats is expected, since the nuclear envelope is an extension of the rough ER (23-26).

TABLE Vi11
Phosphohydrolase activities in liver homogenates and isolated nuclei before correction for the contribution of enzyme in "leaky" membranes Assay media contained the following in a volume of 1 ml: 50 mM Trislcacodylate buffer, pH = 6.5, 10 mg of bovine plasma albumin (defatted), 10 mM phosphate substrates, and appropriate amounts of homogenate or nuclear suspension to measure initial rates. Other details are described under "Experimental Procedures." Glucose-6-phosphatase in the Nuclear Envelope 6-phosphatase activities in both intact and fully disrupted nuclear membranes were virtually identical with their microsomal counterparts. The comparisons include the Michaelis constant for glucose-6-P, the quantitative and qualitative effects of exposures to mannose-6-P, NEM, and protease, the latency of carbamyl-P:glucose phosphotransferase, and the patterns of response of activity and change in latency of glucose-6-phosphatase induced by fasting, diabetes, and cortisol injection.
These findings bear directly and decisively on the questions that have been raised concerning the relevance of the characteristics of glucose-6-phosphatase in intact microsomal vesicles to an understanding of the form and the function of the enzyme as it occurs in uiuo (13)(14)(15)(16). The concept that glucose-6-phosphatase is artificially constrained in intact microsomal vesicles originated (13) as an attempt to explain the phenomenon of latency which is characteristic of the various phosphohydrolase and phosphotransferase activities catalyzed by the enzyme. With the exception of activities with glucose-6-P and the PP, phosphohydrolase, all activities catalyzed by glucose-6-phosphatase of fully disrupted preparations are either highly or completely suppressed in intact microsomes (4,7,8). A position similar to Nordlie and his coworkers has been taken by Wishart and Fry (28), who have suggested that the latency of UDP-glucuronosyltransferase (EC 2.4.1.17) observed in microsomes, but not isolated nuclei, is largely a preparative artifact resulting from vesicularization of the ER membrane.
The argument that latency of these enzymes in microsomes results from "morphological constraints" imposed by the membrane fails to consider the role of the ER membrane as a permeability barrier (8,22,30,(38)(39)(40). Selective permeability is a basic tenet of the substrate transport model of glucose-6-phosphatase, which explains latency in terms of the specificity of the transporters, TI and TS, and the kinetics of the transport-dependent system (2, 4-6, 8, 38-40). The data reported here show clearly that differences observed between untreated preparations of nuclei and microsomes are simply expressions of significant differences in the degree of intactness of their respective permeability barriers.
Since flattened cisternae, characteristic of the ER in situ, are preserved in the intact regions of the envelope of isolate nuclei (23)(24)(25)(26)(27), a comparison of nuclei and microsomes provides a direct unambiguous test of whether fragmentation and vesicularization of the ER membrane artificially alters glucose-6-phosphatase. Thus, the demonstration that the characteristics of the glucose-6-phosphatase system in intact membranes of these preparations, in fact, are identical constitutes the most persuasive evidence to date that the properties of glucose-6-phosphatase in situ are faithfully reproduced with intact microsomes. We find no merit in the arguments (13-16) that the enzyme or translocases of the glucose-6-phosphatase system are artificially constrained in intact microsomes.
It remains our view (Refs. 4 and 7 and see the "Introduction") that glucose-6-P hydrolysis is the only significant physiologic function of glucose-6-phosphatase.