The Binding Protein’s Binding Protein— Clinical Applications of Acid-Labile Subunit (ALS) Measurement1

Robert C. Baxter

Kolling Institute of Medical Research University of Sydney Royal North Shore Hospital St Leonards, NSW 2065, Australia

Address correspondence and requests for reprints to: Robert C. Baxter, Kolling Institute of Medical Research, University of Sydney, Royal North Shore Hospital, St. Leonards NSW, Australia 2065.


    Introduction
 Top
 Introduction
 References
 
More than two decades have passed since Zapf and colleagues (1) first described how "nonsuppressible insulin-like activity" associates with proteins of high molecular weight in serum. Subsequent studies in several laboratories showed that similar high molecular weight forms of somatomedin-like peptides were both GH-dependent and acid-dissociable, leading to the concept that oligomeric protein complexes containing somatomedins included an acid-labile component. The anabolic and mitogenic peptides described in these early studies all turned out to be insulin-like growth factor (IGF)-I or -II, and the other members of the trimeric complex have been identified as IGF binding protein-3 (IGFBP-3), a 40- to 45-kDa glycoprotein with high affinity (approaching 1011 L/mol) for IGF-I and IGF-II, and the acid-labile subunit (ALS), another glycoprotein of 84–86 kDa, which is, as the name implies, irreversibly denatured upon acidification to about pH 4 or below.

ALS is a member of the "leucine-rich repeat" family of proteins; almost 25% of its amino acid content is leucine, predominantly found in characteristic 24-residue repeating units containing 6 leucine residues each. These are predicted to confer marked constraints upon the conformation of the protein, and preliminary molecular modeling in the author’s laboratory suggests a donut-shaped molecule with one surface bearing a considerable negative charge (J. Janosi, unpublished). A carboxy-terminal domain of IGFBP-3 with a concentration of basic residues is known to be involved in the interaction with ALS, and ALS-containing ternary complexes are exquisitely sensitive to dissociation by high pH or ionic strength.

A radioimmunoassay (RIA) for human ALS was described in 1990 (2). Other methods that have been used for semiquantitation of ALS include immunoblot analysis after SDS-PAGE (3) and a binding assay using a covalent complex of iodinated IGF-I and IGFBP-3 (4)–the latter presumed to measure only ALS that is not in ternary complexes ("free" ALS). The presence of ALS has also been inferred in many studies by the observation of IGFs or IGFBP-3 in high molecular weight forms. In such studies, samples must be analyzed by a relatively nondissociating method such as gel permeation chromatography near neutral pH. Gel chromatography under acidic conditions (which denatures ALS), or in the presence of very high salt concentration or high pH (both of which tend to dissociate ALS) is incapable of revealing intact ternary complexes. Similarly, SDS-PAGE cannot be used to demonstrate these complexes unless they are first treated with a covalent cross-linking agent. Under these circumstances, full ternary complexes may be observed, but artifactual, or extremely low affinity, binary complexes of ALS with IGFs or IGFBP-3 may also be stabilized.

The published RIA method measures total serum ALS, and immunodepletion of IGFBP-3-containing complexes from human serum suggests that over half of total ALS was associated with IGFBP-3; the remainder was assumed to be free (2). Estimation of typical molar concentrations of IGFBP-3 (~4 mg/L, 100 nM) and ALS (~25 mg/L, 300 nM) in healthy adult serum using published RIA methods suggest a higher proportion of ALS in the uncomplexed form–up to two thirds. However, a recent restandardization of the author’s assay, using a new ALS preparation calibrated by quantitative amino acid analysis, indicates that midrange ALS values in adult serum are somewhat lower than previously estimated–approximately 250 nM. This still indicates a considerable molar excess of ALS over IGFBP-3, with about 60% of ALS predicted to be in the free form.

In this issue of JCEM, Khosravi et al. (5) (see page 3944) describe a new ELISA for ALS. This method differs from the RIA in several respects. First, whereas the RIA appears to only detect ALS in its native conformation, the ELISA detects ALS with 200-fold increased sensitivity if the protein is first denatured by acidification or SDS treatment. This may result from the fact that the capture and detection antibodies are raised against synthetic amino-terminal and carboxy-terminal ALS peptides, respectively, which presumably have conformations different from those of the corresponding sequences in the native protein.

Second, the authors claim that the ELISA can be used to differentiate free from total ALS. The free ALS assay is broadly similar to that for total ALS, but omits the addition of SDS, resulting in greatly decreased sensitivity. Comparison of free and total ALS levels in sera from 41 adults indicates a weak correlation (r = 0.52) and a slope of 0.47; that is, on average about 50% of total ALS is detectable in the free assay, a result broadly in agreement with previous estimates by other methods. Similarly, approximately 50% of activity in the total ALS assay remains after IGFBP-3-containing complexes are removed from serum by immunoaffinity chromatography, whereas none of the activity measurable in the free assay is removed by this method. It should be possible, using these methods, to show that the fraction of ALS measurable in the free assay has a lower average molecular weight (~85 kDa), when analyzed under nondenaturing conditions on a size-fractionation column, than complexed ALS (~150 kDa), but this demonstration is not included in the paper. It must be commented that the free ALS assay appears to depend on the ALS in serum samples being entirely in the native state, as denaturation vastly increases its apparent immunoreactivity. Whether incorrect sample handling can partially denature serum ALS, leading to a distortion of the free values, remains to be demonstrated.

The available assay methods, in particular the quantitative RIA, have provided the opportunity to learn a great deal about ALS physiology over the past few years. GH appears to be the dominant hormonal regulator of ALS, and exogenous GH increases ALS gene transcription, steady-state messenger RNA (mRNA) levels, output of ALS by hepatocytes, and serum levels of ALS. Exogenous IGF-I has no effect on ALS production by isolated rat hepatocytes, but when administered to healthy humans, suppresses serum ALS levels (6). This is assumed to occur as a consequence of GH suppression by IGF-I, as it is not seen when GH is coadministered (6). IGF-I treatment of patients with GH insensitivity fails to suppress ALS (7), similarly suggesting the involvement of GH in the effect.

While it is established that children with GH deficiency have low serum ALS levels, no published studies have yet compared values in GH-deficient and healthy children with careful age-matching–a comparison that clearly needs to be performed in a number of laboratories to assess the usefulness of ALS measurement in childhood GH-deficiency. GH-deficient adults show variably low levels. In the study of de Boer et al. (8), only 4 of 46 young adults (9%) with GH deficiency of childhood onset had ALS levels above the lower normal limit. Thorén et al. (9) similarly found ALS levels considerably below the normal range in untreated adults with GH deficiency. In contrast, Hoffman et al. (10) reported ALS levels within the normal range in 15 of 22 GH-deficient subjects (68%) of varying ages and etiology of GH deficiency—a result similar to that presented by Khosravi et al. in this issue (5). The discrepancies between these studies may, to a large extent, reflect different study populations, adults with childhood-onset GH deficiency showing lower ALS values (and lower IGF-I and IGFBP-3 values) than those with GH deficiency of adult onset. Clearly, a normal ALS, IGF-I, or IGFBP-3 level in an adult does not provide conclusive evidence of normal GH secretory status. There is, nevertheless, a significant relationship between ALS and IGF-I levels and the daily GH dose in adults receiving GH replacement; IGFBP-3 levels, in contrast, do not show this relationship (9).

ALS levels are significantly elevated in acromegalic subjects (2, 10). Hoffman et al. (10) have reported mean levels in acromegalic subjects 2.5-fold above the normal mean, with 20 out of 22 subjects (91%) showing values above the normal range. In contrast, Khosravi et al. (5) reports ALS levels in 4 out of 12 acromegalic subjects (33%) overlapping with the normal group. De Boer et al. (8) have also assessed the potential use of ALS measurement in monitoring a state of GH excess, in a study evaluating the optimal replacement dose of GH in GH-deficient adults. This study concluded that ALS measurement was not as sensitive a marker as IGF-I, because upon treatment with a GH dose that elevated IGF-I levels beyond the upper limit of normal, ALS levels, like IGFBP-3 levels, remained within the normal range. It must be noted, however, that such an evaluation requires accurately assessed normal ranges for healthy populations. Such ranges have been more extensively established and validated for IGF-I than for either IGFBP-3 or ALS. These data do, however, support the conclusion that an elevated ALS level is good prima facie evidence of GH excess.

Human serum ALS has been studied in a wide variety of other physiological and pathological conditions. ALS levels are lowest in neonates, rise steadily to peak values in mid- to late-puberty, and show a slow decline throughout adult life (2). In pregnancy, ALS levels increase with advancing gestation, and the ALS:IGFBP-3 ratio also shows a significant increase (11). Because ALS may help to stabilize ternary complexes affected by the action of pregnancy-specific IGFBP-3 proteases, the increasing ALS:IGFBP-3 ratio through gestation may be an important factor in maintaining the high serum IGF levels characteristic of advancing pregnancy.

Patients with nonislet cell tumor hypoglycemia (NICTH) often have considerably decreased ALS levels, probably resulting from suppressed GH secretion. This is a comparable situation, arising from an IGF-II-secreting tumor, to that seen following IGF-I administration. Both GH and corticosteroid administration have been reported to increase ALS levels and restore normoglycemia (12). The effect of corticosteroid is paradoxical, as studies in vitro show a suppressive effect of dexamethasone on ALS production and mRNA levels. A marked concordance between serum ALS and glucose levels has been described in a patient with NICTH over a prolonged period of treatment (12). This is consistent with the concept that ALS may play a key role in glucoregulation by stabilizing IGFs in ternary complexes, thus preventing their egress from the circulation to the tissues, where they could exert insulin-like activity.

ALS is also subject to nutritional regulation, and in several animal studies serum ALS levels have been shown to decrease after fasting or caloric restriction. The nutritional effect may be mediated by changes in cyclic AMP levels, as cyclic AMP has recently been shown to down-regulate hepatic ALS mRNA levels and inhibit ALS secretion (13). Critically ill patients have markedly decreased serum ALS levels, which correlate strongly with levels of the conventional nutritional indices transferrin, retinol binding protein, and particularly prealbumin (14). Indeed, ALS levels showed a closer association with these proteins than levels of either IGF-I or IGFBP-3. This suggests that further evaluation of ALS as an indicator of metabolic status in the critically ill is warranted. In these patients there appears to be a coordinated response in IGFs, IGFBP-3, and ALS over the recovery period. In particular, IGFBP-3 and ALS levels remain closely associated, an observation also made in other clinical conditions and confirmed in the study of Khosravi et al. (5).

In patients with other potentially catabolic conditions, including insulin-dependent diabetes and severe burns, ALS levels are also significantly decreased, and in both cases respond to insulin therapy (15). Finally, in patients with hepatic cirrhosis, ALS levels have been shown to decrease with increasing severity of disease, with values in Childs Pugh group C patients 80% lower than those in Childs Pugh group A patients (16). ALS measurement might therefore have potential as a marker of worsening hepatic function. In contrast, serum IGF-I levels in the same study, though lower in cirrhotic patients than in healthy controls, showed no discrimination between patients in Childs Pugh groups A–C (16).

What are the potential clinical uses of ALS assays? This question might also be asked in a different form: in what situations will ALS measurement provide more useful information than IGF-I? The answer must be cautious at present, as the majority of clinical studies on ALS have involved measurement in a single laboratory using a single RIA method. It may be envisaged that, of the assays that may be introduced over the next few years, different methods could yield slightly different results because they are affected to varying degrees by the conformation of the ALS in serum and by the fraction bound in ternary complexes. Furthermore, ALS is heavily glycosylated, and some antibodies will undoubtedly detect carbohydrate epitopes. Whether ALS, like other hepatic proteins, shows microheterogeneity in its glycosylation, dependent on nutritional or disease states, will no doubt form the basis of future studies.

Despite these qualifications, is appears that ALS is a valuable index of GH status in many situations and could have useful applications as a marker of nutritional status in critical illness and of hepatic function in cirrhosis. ALS offers the potential advantage over IGF-I that its measurement is unaffected by IGFBPs, which can interfere in many conventional IGF-I assays. Although the protein is quite difficult to work with in the purified form, ALS levels in serum samples appear to be very stable to a variety of storage conditions and to several freeze-thaws. As a protein with an almost exclusively hepatic origin, it may reflect hepatic GH responsiveness better than IGF-I, the production of which is stimulated by GH in many tissues. The question of the potential value of measuring free ALS, in addition to, or as an alternative to, total ALS, remains unanswered, as the available data are as yet insufficient to suggest clinical situations where free ALS levels may offer a particular diagnostic advantage.

Perhaps the strongest conclusion that can be drawn at this stage is that it is too early to draw a conclusion. To date, there has been insufficient evaluation of the diagnostic applications of ALS measurement in many centers to allow the extent of its utility, and its limitations, to be well understood. There is likely to be a significant improvement in this situation soon, as several commercial ALS assays are currently being released. In this context, the ELISA method reported by Khosravi et al. (5) is a welcome addition. It will be necessary to test these assays in a wide variety of clinical conditions to allow an accumulation of sufficient information on which to make a final judgement. In the meantime, there is no doubt that ALS measurement is providing valuable information about the physiology and pathology of the GH-IGF system.


    Footnotes
 
1 The author’s studies are supported by the National Health and Medical Research Council, Australia, and by the Cooperative Research Centre for Diagnostic Technologies, Australia. Back

Received October 2, 1997.

Accepted October 3, 1997.


    References
 Top
 Introduction
 References
 

  1. Zapf J, Waldvogel M, Froesch ER. 1975 Binding of non-suppressible insulin-like activity to human serum. Evidence for a carrier protein. Arch Biochem Biophys. 168:638–645.[Medline]
  2. Baxter RC. 1990 Circulating levels and molecular distribution of the acid-labile ({alpha}) subunit of the high molecular weight insulin-like growth factor-binding protein complex. J Clin Endocrinol Metab. 70:1347–1353.[Abstract]
  3. Liu F, Hintz RL, Khare A, DiAugustine RP, Powell DP, Lee PDK. 1995 Immunoblot studies of the IGF-related acid-labile subunit. J Clin Endocrinol Metab. 79:1883–1886.[Abstract]
  4. Baxter RC. 1988 Characterization of the acid-labile subunit of the growth hormone-dependent insulin-like growth factor binding protein complex. J Clin Endocrinol Metab. 67:265–272.[Abstract]
  5. Khosravi MJ, Diamandi A, Mistry J, Krishna RG, Khare A. 1997 Acid-labile subunit of human insulin-like growth factor binding protein complex: Measurement, molecular and clinical evaluation. J Clin Endocrinol Metab. 82:3944–3951.[Abstract/Free Full Text]
  6. Kupfer SR, Underwood LE, Baxter RC, Clemmons DR. 1993 Enhancement of the anabolic effects of growth hormone and insulin-like growth factor I by use of both agents simultaneously. J Clin Invest. 91:391–396.[Medline]
  7. Gargosky SE, Wilson KF, Fielder PJ et al. 1993 The composition and distribution of insulin-like growth factors (IGFs) and IGF-binding proteins (IGFBPs) in the serum of growth hormone receptor-deficient patients: Effects of IGF-I therapy on IGFBP-3. J Clin Endocrinol Metab. 77:1683–1689.[Abstract]
  8. De Boer H, Blok GJ, Popp-Snijders C, Stuurman L, Baxter RC, Van der Veen EA. 1996 Monitoring of growth hormone replacement therapy in adults, based on measurement of serum markers. J Clin Endocrinol Metab. 81:1371–1377.[Abstract]
  9. Thorén M, Hilding A, Baxter RC, Degerblad M, Wivall-Helleryd IL, Hall K. 1997 Serum insulin-like growth factor I (IGF-I), IGF-binding protein-1 and -3, and the acid-labile subunit as serum markers of body composition during growth hormone (GH) therapy in adults with GH deficiency. J Clin Endocrinol Metab. 82:223–228.[Abstract/Free Full Text]
  10. Hoffman D, Baxter R, O’Sullivan A, Crampton L, Ho K. 1997 Serum acid-labile subunit in adult growth hormone deficiency and acromegaly. Endocrinol Metab. 4 (Suppl A):23 (Abstract).
  11. Suikkari A-M, Baxter RC. 1992 Insulin-like growth factor binding protein-3 is functionally normal in pregnancy serum. J Clin Endocrinol Metab. 74:177–183.[Abstract]
  12. Baxter RC, Holman SR, Corbould A, Stranks S, Ho PJ, Braund W. 1995 Regulation of the insulin-like growth factors and their binding proteins by glucocorticoid and growth hormone in nonislet cell tumor hypoglycemia. J Clin Endocrinol Metab. 80:2700–2708.[Abstract]
  13. Delhanty PJD, Baxter RC. The regulation of acid-labile subunit gene expression and secretion by cyclic AMP. Endocrinology. In press.
  14. Baxter RC, Hawker FH, To C, Stewart PM, Holman SR. 1997 Thirty-day monitoring of insulin-like growth factors and their binding proteins in intensive care unit patients. Growth Reg. 7(2):1–11
  15. Bereket A, Wilson TA, Blethen SL, et al. 1996 Regulation of the acid-labile subunit of the insulin-like growth factor ternary complex in patients with insulin-dependent diabetes mellitus and severe burns. Clin Endocrinol (Oxf). 44:525–532.[Medline]
  16. Donaghy AJ, Baxter RC. 1996 Insulin-like growth factor bioactivity and its modification in growth hormone resistant states. Bailliere’s Clin Endocrinol Metab. 10:421–446.[Medline]