Journal of Immunodeficiency & DisordersISSN: 2324-853X

All submissions of the EM system will be redirected to Online Manuscript Submission System. Authors are requested to submit articles directly to Online Manuscript Submission System of respective journal.

Review Article, J Immunodefic Disor Vol: 3 Issue: 1

Alpha-Fetoprotein as a Biomarker in Immunodeficiency Diseases: Relevance to Ataxia Telangiectasia and Related Disorders, Immunodiagnostics

Gerald J Mizejewski*
Wadsworth Center, Division of Translational Medicine, New York State Department of Health, The Empire State Plaza, Albany, NY 12209, USA
Corresponding author : Gerald J Mizejewski
Wadsworth Center, New York State Department of Health, The Empire State Plaza, PO Box 509
Tel: 518 486-5900; Fax: 518 402-5002
E-mail: [email protected]
Received: October 30, 2013 Accepted: March 17, 2014 Published: March 24, 2013
Citation: Mizejewski GJ(2014) Alpha-Fetoprotein as a Biomarker in Immunodeficiency Diseases: Relevance to Ataxia Telangiectasia and Related Disorders. J Immunodefic Disor 3:1. doi:10.4172/2324-853X.1000108


 Alpha-Fetoprotein as a Biomarker in Immunodeficiency Diseases: Relevance to Ataxia Telangiectasia and Related Disorders

Human Alpha-fetoprotein (AFP) involvement as a biomarker in immunodeficiency disorders (IDD) has received little attention in the biomedical literature due to the recent focus on AFP epitope analysis, vaccine development, and regulation of T- and B-cell immune responses. Thus, the literature is currently devoid of a review which encompasses the use of AFP as a biomarker in the different types of IDDs. In order to update and increase present knowledge regarding an AFP-IDD relationship, the present review surveys and addresses the presence of AFP serum levels in patients with IDDs such as acquired immunodeficiency syndrome (AIDS), severe combined immunodeficiency (SCID), adenosine deaminase deficiency (ADD), and in pregnant women infected with the human immunodeficiency virus (HIV).

Keywords: Alpha-fetoprotein; Immunodeficiency; HIV/AIDS; Ataxia; Lymphoreticular; Chemokines; DNA Repair; Chromosome; Cancers


Alpha-fetoprotein; Immunodeficiency; HIV/AIDS; Ataxia; Lymphoreticular; Chemokines; DNA Repair; Chromosome; Cancers


Human alpha-fetoprotein (HAFP) is a tumor-associated fetal glycoprotein (oncofetal protein) that is associated with certain malignant neoplasms and with fetal defects during ontogenic growth and differentiation [1]. In the clinic, AFP has served as a serum biomarker for adult liver diseases such as alcoholic cirrhosis, necrosis, viral hepatitis, hepatocellular carcinoma, and hepatoblastoma [2]. It also occurs in other cancers including teratomas, testicular and yolk sac tumors, gastric, pancreatic, and esophageal malignancies. During pregnancy, AFP plays a role in development of the embryo and fetus and has been employed as a biomarker in fetal, prenatal, and pediatric disorders such as Beckwith-Wiedemann syndrome, Fanconi’s anemia, and pancreatic blastomas [3,4]. In addition to the IDDs mentioned above, ataxia telangiectasia (AT) is an immunodeficiencyrelated disorder involving both the immune and nervous systems that is associated with increased AFP serum levels in both juveniles and adults [5,6].
Alpha-fetoprotein consists of a single polypeptide chain composed of 609 amino acids comprising a glycoprotein of 69 kilodaltons (Figure 1). HAFP belongs to the albuminoid gene family which is structurally characterized by cysteine residues that are folded into layers that form loops dictated by disulfide bridging [7,8]. A “hinge concept” was derived from the observation that HAFP has two fewer cysteines and one less disulfide bridge than human albumin, providing it with a means of polypeptide chain flexibility (Figure 1; Panel-D). The lack of a disulfide bridge in the latter half of domain-2 (amino acids residues #295 to 346) results in the loss of a rigid link to domain-3 producing a flexible hinge between domains-2 and -3 [9]. In comparison to the rigidity of human albumin, the proposed hinge region would permit a greater rotational flexibility for domain-3. Such a hinge might permit rotational flexibility of the polypeptide backbone of AFP providing an advantage for proteinto- protein interaction and possibly receptor docking [10]. Receptor and ligand binding to HAFP is known to occur on domain-3 which is an extended arm of the U- or V-shaped shaped AFP molecule. Thus, regarding domain functional assignments (Figure 1, Panel-E), domain-3 is involved with receptor binding, heptad and dimerization sites, estrogen (steroid), kinesin, and retinoic acid binding. See Figure 1, Panel-E for the AFP domain-1 & 2 functional assignments confirmed and proposed to date.
Figure 1: The three-domain structure of human alpha-fetoprotein (HAFP) is displayed as three differenct in panels A-C. The disulfide bridging by 32 cysteine residues is shown in a linear extended chain model; in Panel D, the disulfide bridging dictates the formation of the poly-peptide chain loops. Panel A: A three-dimensional helix/ribbon model of the AFP molecule is depicted. Note the overall U or V-shape that the 3-D structure assumes. The 3-D model was derived from Reference [10,49]. Panel B: An electron dot contour model is shown which was produced by a combination of electron microscopy, image processing, and circular dichroism procedures [8]. Panel-C: The structure of AFP is demonstrated as a U or V-shaped stick chain configuration. Note that the hinge region is produced as a result of lacking one disulfide bridge between the second and the third domain (Panel D). This feature bestows a rotational flexibility to the third domain in contrast to domain one and two, providing an advantage in protein-to-protein (ligand) interaction and docking. Observe that the third domain rectangular blocks among the three panels are connected to illustrate that ligand-to-AFP binding can readily occur in the extended arm of the V-shaped molecule. Panel-D: The amino acid single letter code is displayed. Panel-E: The listing illustrates functions that have been proposed and assinged to the three domains. This figure represents a reconfigured diagram redrawn with permission from a copyrighted figure derived from: Reference [7], 2004, Exp. Biol. & Med., Society of Experimental Biology & Medicine. References [8,9] were used as informational resources.
The present review and update is based on previous studies showing that HAFP is involved in the regulation of immune responses, both the innate and acquired types of immunity. Indeed, a multitude of reports over many years have demonstrated that human AFP associated with immunoenhancement and/or immunosuppression involving both T- and β- lymphocytes, dendritic cells, helper T-cells, thymocytes and cytotoxic lymphocytes [1,2]. As discussed below, AFP levels have been ascribed to immunodeficiency disorders such as AIDS, SCID, adenosine deaminases (ADA) deficiency and ataxia telangiectasia (AT) [11]. To date, the involvement of AFP in some of these disorders has been proposed and studied but there is a lack of co-ordinated efforts to review these findings, sparse as they may be. Thereby, the objectives and goals of the present treatise are four-fold. First, the AFP serum levels in these various IDDs will be surveyed and discussed in light of their prevalence and clinical significance. Second, the biological aspects of AFP in these ID disorders will be reviewed, updated, and discussed. Third, a proteomic approach to elucidate protein-to-protein interactions of AFP with some possible protein effectors of these disorders will be presented with the aid of computer methodology. The proteomic results from computer modeling, protein-to-protein interaction software, and third domain-structure analysis will be combined in an attempt to identify IDD- related protein interaction sites on the AFP molecule. Finally, based on computer modeling localization of the AFP-IDD interaction peptide sites, AFP involvement in these disorders will be discussed.

Adenosine Deaminase Deficiency

Adenosine deaminase (AD) is an enzyme involved in purine metabolism, in procuring adenosine from foodstuffs, in nucleic acid turnover, and in maitaining the immune system. Adenosine deaminase deficiency (ADD) is an autosomal recessive inherited disorder that damages the immune system resulting in immunodeficiency and severe combined immunodeficiency disease [11]. ADD occurs in about one in 100,000 live births worldwide and can manifest in infancy, childhood, adolescence, or adulthood [12]. Patients with SCID virtually lack immune protection against bacteria, viruses, and fungal infections which normally exist in non-affected patients. One of the metabolic deficiencies of ADD is due to a lack of the enzyme adenosine deaminase, a condition which is toxic to cells (especially erythrocytes and lymphocytes) due to inhibition of ribonucleotide reductase and subsequent DNA synthesis. The inhibition of the AD enzyme prevents mitosis in cells susceptible to such conditions. ADD is further involved in the purine salvage pathway and results in increased S-adenosylhomocysteine (SAH) levels. This results in inactivation of SAH hydrolase which causes impairment of normal methylation reactions. Lack of these enzymes affects immature lymphocytes which fail to differentiate and mature into end-stage cells. For example, T-cells undergoing proliferation and development in the thymus are typically affected resulting in a small, underdeveloped thymus organ. As a result, the T-cell branch of the immune system is severely compromised or entirely lacking. Infants with ADD and SCID usually experience disabling conditions such as chronic diarrhea, widespread skin rashes, growth retardation, and pneumonia [13]. Some patients also display neurological symptoms such as developmental delay, movement disorders, and hearing loss (see below). Patients with ADD further exhibit recurring upper respiratory and ear infections and later develop chronic lung damage, allergies, and a variety of health problems.
The gene for AFP is independently modulated in fetal liver during development, while postnatal AFP gene expression is normally silenced by a process of DNA methylation. A previous report indicated that the induction of AFP synthesis in teratoma cells undergoing differentiation occurs after a genome-wide loss of DNA methylation [14]. Another study demonstrated that deoxyadenosine is a suppressor of methylation reactions. These studies promoted investigations to measure AFP levels in patients exhibiting a variety of immunodeficiency and ADD-related disorders. Overall, patients with such immunodeficiency disorders are unable to metabolize deoxyadenosine in a normal fashion [15]. Immunodeficiency disorders were studied in afflicted patients and those displaying normal levels of adenosine deaminase (Table 1). The study further included non-immunodeficient patients who were receiving parenteral hyperalimentation.
Table 1: Human Alpha-Fetoprotein (AFP) levels in various immunodeficiency diseases including Ataxia Telangiectasia and Oculomotor Apraxia.
As shown in Table 1, all six patients with adenosine deaminase deficiency displayed elevated levels of serum AFP with levels found in disorders such as AT, while some patients demonstrated levels reported in malignant teratomas and hepatomas (Figure 2). In 24 patients with immunodeficient and normal adenosine deaminase levels, seven displayed raised AFP levels. Four patients were studied in order to determine the effect of liver disease and parenteral hyperalimentation on AFP levels; 2 such patients demonstrated elevated levels of AFP. When immunodeficient patients were parsed into those with either liver disease or parenteral hyperalimentation, no correlations with AFP serum levels were found. Such results suggested that elevated AFP levels (of hepatic origin) need not be specific to the immunodeficiency aspect of these disorders; moreover, the cause of elevated AFP levels in patients with related immunodeficiency disorders has yet to be determined.
Figure 2: Serum alpha-fetoprotein (AFP) levels (ng/mL) were measured by radioimmunoassay in normal adults and in patients with various liver disorders, combined immunodeficiency disorders (CID), ataxia telangiectasia (ATT), oculomotor apraxia (OMA), and various malignant tumors. Serum AFP levels of liver cirrhosis, hepatitis, and necrosis patients (left side of Figure) were compared to those with AT, OMA, HIV, CID, teratomas and hepatomas in the middle and right side of the Figure. AFP levels in HIV pregnant women ranged from 24 to 150 ng/mL. The total number (N) of samples assayed in each column is shown below each listed disorder. Samples measured less than 20 ng/mL are displayed by the number in the rectangular block above the named disorder. Each dot shown in the vertical column plots indicates one patient. Data for the Figure was extracted from References [1,2,4,22,116,117].

AFP and Acquired Immunodeficiency Disorders (AIDS; HIV)

As increasing number of anti-retroviral drugs are now being administered to pregnant women infected with the human immunodeficiency virus (HIV), either for the prevention of vertical transmission (mother-to-fetus) or as therapy to the mother [16-20]. However, such drugs can carry risks as well as benefits for both the fetus and neonate, depending on the individual’s perinatal drug metabolism. Some drugs given to the neonate can accumulate in the eyes, larynx, tongue, and digestive tract. Despite the possible vertical transmission of HIV associated with pregnancy and the neonatal period, more than 90% of the instances of HIV transmission have been found to occur in the postnatal/neonatal stages of the infant rather than during pregnancy. Maternal-infant transmission of HIV usually occurs through vaginal secretions during labor, breast feeding, or through exposure of the fetus to maternal blood [19,20]. In fact, AIDS symptoms are delayed in the infected infants, with HIV detection occurring later in the infant period [21,22]. Even though there may be several interpretations of the lack of fetal infection in HIV-positive mothers in the first and second trimesters [16-22], a possible connection to the presence of HAFP can be discussed. Previous studies suggested that the relatively low transmission rate of HIV-1 to the fetus during pregnancy may be associated with raised maternal serum AFP levels [16-22].
Uriel and his associates working in the 1987-89 period demonstrated that the specific uptake of HAFP occurs via receptors on human T-lymphoblast cells during antigen-induced transformation and in malignant lymphoid cells [23,24]. As a follow-up of these studies, Uriel’s group reported an impairment in the ability of AIDS (HIV) patients’ peripheral blood mononuclear cells to internalize AFP [25,26]. An AFP endocytosis assay clearly revealed a defective uptake of AFP in AIDS patients, both in lymphoadenopathy syndrome and in mitogen-responsive T cells of asymptomatic patients. At that time, the authors proposed that the reduced capacity to bind and internalize AFP in early-stage nonsymptomatic HIV/AIDS patients provided the basis for a possible prognostic test [26]. It was further reported that a defective AFP cell receptor uptake occurred with a decreased expression of IL-2 receptors on the lymphoid cell surface [27]. Thus, the investigators noted that the seroconversion from HIV-negative to HIV-positive correlated with the onset of a progressive deterioration of the patients’ AFP receptor uptake capability. Thus, the presence of AFP at the cell surface of monocytes from AIDS patients may in some way interfere with the process of HIV-cell membrane fusion and subsequent transfection.
Uriel’s team then sought to determine whether the AFP-uptake impairment was due to the presence of an AFP receptor blockade or to a target signal transduction defect. They studied AFP endocytosis in peripheral mononuclear cells (PMCs) from asymptomatic HIVpositive patients as well as in HIV-transfected PMCs (in vitro) from healthy donors in phytohemagglutinin-stimulated and nonstimulated conditions. Their results revealed that the defective AFP endocytosis was a consequence of an abnormal mitogenic response of PMCs associated with the presence of the HIV virus [26]. Thus, their findings included both the state of the T-cell activation associated with HIV infection and the PMCs’ responsiveness to mitogenic stimulation. Uriel and co-workers further reported a considerable loss of membrane fluidity of the PMCs in AIDS patients, as evidenced by an elevated cholesterol/phospholipid (CH/PL) ratio in cell membranes from AIDS patients [27]. Relative to normal cells, the expression of AFP and other receptors appeared considerably reduced in AIDSrelated complex disorders and in AIDS patients. In summary, the HIV infection disrupted the fluidity of the cell membrane and altered the normal sequences of lymphocyte antigen-activation and blast-cell transformation events. In addition, the carbohydrate content of the receptor and the ligands may also play a role [28,29] and might have influenced AFP interactions with HIV-1 co-receptors [30,31].
The HIV-1 virus penetrates the CD4+ T-cell membrane by means of a cascade of molecular interactions (fusions) between the virion envelope glycoprotein (ENVG) and participation of at least two specific cell-surface receptors [32-35] (Figure 3). The gp120 virus ENVG first binds to the CD4 receptor and induces a conformational change in gp120, revealing an amino acid sequence involved in virus binding to the T-cell CCR-5 chemokine co-receptor [26-29]. Binding to the co-receptor further triggers a conformational change in the gp41-virus ENVG which leads to fusion of the virus envelope with the CD4+ cell membrane. The fusion/entry reactions present multiple protein targets for therapy since the CCR5 chemokine co-receptors are G protein-coupled signal-transducing agents. It has been shown that AFP can compete for binding to these co-receptors (see below). It was then proposed that AFP could stereochemically interfere with fusion of the gp120 ENVG to the CD4 + cell surface membrane, thereby impairing viral transmission/infection during pregnancy [30,31].
Figure 3: The HIV-entry process in host cells (T-cells/monocytes) consists of a series of coordinated interactions; namely, 1) binding to two or three different cell membrane receptors (CD4 and CXCR4 or co-receptor CCR5) and 2) fusion of the cell membrane with viral coat glycoproteins (gp120 and gp41). The total viral coat glycoprotein is a trimer composed of two gp120 subunits and one gp41 unit. The initial binding occurs between gp120 and the CD4 receptor, after which the gp120 undergoes a conformational change allowing further binding to CCR5 and/or CXCR4. It is believed that binding to the CD4 and chemokine receptors removes a steric hindrance allowing the gp41 segment to mediate cell membrane fusion and subsequent viral entry. The figure displays the HIV-1 virus particle approaching and positioning itself above the T-cell or monocyte cell surface. Overall, the figure demonstrates virus-membrane fusion and shows that AFP derived from the extracellular fluid (ECF) may compete with binding of the gp120 V3C loop clade to CD4 receptor; this event might interfere with viral-to-cell membrane fusion (see text) and CCR5 coreceptor interaction. References [30,31,118] were used as informational resources.
In a 1997 study, investigators reported that AFP interacted in vitro with the HIV-1 gp120 viral coat protein [35]; thus, AFP appears capable of blocking the infection of primary monocyte-derived macrophages by various HIV-1 viral strains. Serving as a steric hindrance agent, AFP could act during CD4-independent stages of virus fusion to the cell membrane of macrophages. In that report, AFP was shown to block the binding of HIV specifically at the virus V3 loop clade consensus peptide, thus interfering with viral post-binding events during the HIV-1 infection process of primary macrophages [31]. In addition, a carbohydrate-involved HIV infection pathway was found, which depended on the host macrophage cell type and on differences in the glycan structure of the specific co-receptors involved in HIV cell entry [29]. However, it has not been shown whether AFP is involved at the glycan-associated level. Nonetheless, it was demonstrated that HAFP specifically interacted at the primary macrophage cell surface and competed with the gp120 V3C for binding of HIV-1 to these cells [30]. In addition, the use of antibodies to the CCR-5 chemokine receptor was shown to inhibit AFP binding to the macrophages. Native intact (but not heat-denatured) AFP specifically bound to electroblotted gp120 V3C- ligand, that is, the CCR5 cell surface receptor. These data demonstrated that an AFP steric blocking effect during HIV infection was directly associated with an AFP-virus interaction concurrent with the HIV infection, and directly involved AFP binding to the CCR5 and other chemokine co-receptors. This set of observations could provide the basis for at least one plausible explanation for the lack of vertical transmission of HIV-1 infection repeatedly observed during pregnancy [18-21].
Studies from the above investigations further revealed that HAFP was able to bind to CCR5 receptors at both high- and lowaffinity binding sites (KA = 5.15 and 100 nM, respectively) localized on monocyte-derived macrophages [31]. The CCR5 chemokine receptor is known to be localized in juxtaposition to the CD4 receptor and serves as a co-receptor (with CD4) for the HIV uptake and transfection process [32-34]. Both protein-to-protein interaction and lectin carbohydrate involvement were established as being involved with the binding process; such experiments utilized treatments such as heat denaturation and neuraminidase exposure to AFP. As alluded to above and in Figure 4, CCR5 ligands (MIP, MCP,Rantes, Eotaxin) were also capable of displacing AFP from its receptor binding to the macrophage cell surface [35]. Finally, it was shown that HAFP could bind to the CCR5 receptor expressed on HeLa cells, but not on HeLa cells lacking CCR5 receptor expression. Finally, further studies revealed that AFP binds not only to the CCR5 receptor associated with CD4 receptors expressed on primary macrophages (monocyte derived), but that CXCR4 could substitute as a co-receptor for CCR5 in HIV-1 infection on T-cells (34) (Figure 3). Thus, the binding of AFP to CCR5 and CXCR4 co-receptors provides additional supporting evidence to explain the apparent lack of vertical transmission to the fetus.
Figure 4: The chemokine gene family is composed of four member groups based on the amino-terminal positioning of their cysteine (C) amino acids (AA). The four family members of chemokine ligand peptides (60-90 AA in length) are divided and subgrouped as follows: Group I has a single cysteine (C) residue in the amino terminus of the molecule; Group II has a CC grouping; Group III has the CXC configuration; and Group IV has the CX3C sequence where X is any AA residue (top portion). The chemokines are the ligands (binding agents) that bind to G-coupled heptahelical receptors; such receptor molecules weave through the cell membrane seven times as a transmembrane (TM) domain. By means of GenBank AA sequence matching, it was observed that HAFP resembles a segment of the chemokine peptide, especially the Group-II (CC) members, GROα, MiP-1B, Eotaxin, Rantes, and MCP-1 chemokine (see subpanels). It is proposed from AA matching (identity/ similarity) evidence that portions of HAFP may pose as a chemokine (fragment) decoy or appear as a chemokine-mimicking protein. Observe the block insets on each panel for the chemokine homolgy comparisons and AFP-matched identities. Note in the figure that the AFP cysteine bridge clusters resemble those of the CC-chemokine ligands. Observe also that CCR5 is the receptor that binds Rantes, Eotaxin, and MIP. Thus, it might be plausible that AFP mimics a cc-chemokine peptide which allows binding to a cc-chemokine receptor, thus influencing the receptor conformational change; in other cases AFP might inhibit cell migration (spreading), adhesion, and proliferation. The Amino-acid sequence matching (identities and similarities) were performed by a FASTA analysis using methods described in Reference [119].

Maternal Serum AFP Levels in Pregnant Women with HIV

It is germane to the present review that maternal serum (MS) AFP levels have been reported in pregnant women with HIV in various investigations [22,36-38] (Table 1). The parameters measured were CD4 counts/percentages and serum viral load. Serum screen results in HIV infected pregnant women by Gross et al. [22] showed elevations in MSAFP levels in HIV-infected pregnant women as compared to controls. Although MSAFP levels correlated in a positive fashion to maternal viral load, CD4 counts and percentages were not significantly altered. Moreover, a trend was present which showed increased MSAFP levels in pregnant AIDS patients compared with pregnant HIV patients (MSAFP MoM = 1.71 versus 1.2 MSAFP MoM, P = 0.067). Following birth, all patients in these studies by Gross et al had normal pregnancy outcomes and the neonates were HIV infection-free up to one year of follow up. Overall, the study by Gross et al showed elevations in both MSAFP and MShCG; however, a study by Yudin et al. [36] demonstrated only MShCG raised levels, while other studies showed neither analyte elevations [37,38]. The study by Gross et al was deemed important to the present review due to: a) a higher blood viral load (3-15 fold higher) than other studies; b) less patients that received anti-retroviral drug therapy (14% versus 48-62%); and c) a higher classification of HIV to AIDS progression in the patients (31% versus 4-6%). For comparative purposes of AFP physiological activity, it was considered crucial that the patients studied had high viral blood loads and most were not receiving antiretroviral drug therapy that could interfere with the interpretive effects of AFP on virus fusion and cell entry (Figure 3). As shown below, drug treatment of HIV-infected women resulted in lowering MSAFP levels.
In their discussion, Gross et al. [22] proposed two hypotheses to explain their findings of elevated MSAFP with increased viral load. Having discounted chronic placental damage as a possible cause of elevated MSAFP levels, the authors suggested increased maternal AFP immunoregulatory activity during progression of the HIV disease. During the last two decades the immunoregulatory role (enhancement/suppression) of human AFP has remained highly controversial with results difficult to duplicate among different laboratories. Unfortunately, the references cited to support their contention of an AFP immunoregulatory role employed largely rodent cells, which were used as in vitro models studied during the 1980- 1990s. However, more recent research has addressed the physiologic and biochemical aspects of AFP interaction with membrane receptor complexes on macrophages and monocytes. Hence, it can be proposed that HAFP binding to the macrophage/monocyte CCR5 receptor at the placental interface might enhance the fetal-to-maternal gradient of AFP placental transport, thus increasing AFP levels in the maternal circulation. It is tempting to speculate that the positive correlation observed between increasing viral load and elevation of MSAFP was related to an AFP stereochemical interference with viral fusion to the CD4 + cell surface receptor and its co-receptors, thus impairing viral transmission and down-regulating the CD4 counts. In this regard, it should be noted that all HIV-infected mothers in the study had normal pregnancy outcomes. None of the screened infants followed for up to one year developed HIV. Therefore, receptor interference AFP may indeed play a role in the lower incidence of HIV vertical transmission from mother to fetus as discussed above [16-17].

MSAFP Levels in Pregnant HIV Women Administered Protease Inhibitor Drugs

The impact of protease inhibitor (PI) therapy on HIV transmission and MSAFP levels in pregnant women was reported in a study by Einstein et al. [39] (Table 1). PIs are thought to decrease placental trophoblastic cell secretion of hormones in vitro and may affect transplacental permeability and function [36]. Einstein’s study reported no differences in initial viral load or CD4 counts between case and control groups. However, significantly lower MSAFP MoM values were found for the women treated with PIs compared with HIV pregnant women who received no anti-retroviral PI treatments (MoM = 0.97 versus 1.2 MoM, P = 04). Interestingly, the Einstein group hypothesized that placental and/or immune dysfunction may have contributed to their reported observations. The Einstein findings of either decreased or unchanged MSAFP levels in pregnant HIV-infected women treated with drugs are in agreement with other studies of this nature [36-38].
An alternate explanation might be directed at the physiological properties of AFP itself. It has been known since the 1980s that human AFP can bind serum and tissue protease inhibitors and small molecule PIs such as arginine, benzamidine, and trayslol [40,41]. One could propose that HAFP might bind but not cleave a protease substrate, such that AFP could serve as a protease mimic or decoy enzyme. Increased serum levels of AFP have been reported after PIs were experimentally stripped from AFP allowing increased quantities of AFP to be measured by immunoassays. Furthermore, HAFP displays amino acid sequence identity with tissue-type plasminogen activator, a serine protease for plasminogen [42]. Inter-alpha trypsin inhibitor (IATI), a Kunitz-type protease inhibitor for trypsin and plasmin is also a binding substrate for chymotrypsin. Interestingly, HAFP was found to compete with chymotrypsin for binding to IATI with a Ka = 5.6 x 10-7M [38]. Hence, PIs administered in the first trimester may have affected AFP function and altered its transplacental passage since fetal serum AFP concentrations peak at 9-10 weeks and decline in the remaining trimesters with a half-life of 1 to 3 days. One could propose that PI administration affected AFP serum concentrations in several ways. First, AFP might bind PIs and modulate feedback regulation of AFP synthesis in the yolk sac and fetal liver. Second, the masking of epitope sites on HAFP, (due to PI binding), might have resulted in reduced antibody binding to AFP and lowered its amounts available for immunoassay detection of the total serum HAFP. Third, an altered half-life of AFP may be the result of an increased vascular clearance of a conformationally modified form of AFP bound to a PI. It has been reported that low MSAFP levels have been associated with growth restriction and small-for-gestational age (SGA) newborns [43]. Since SGA at delivery was reported at 19% in the PI-group and only 4% in the non-PI group, low MSAFP levels might have played a role in the emergence of small-for-gestational age newborns. Whatever the mechanism, lowered MSAFP values reported in the HIV pregnancies would require further studies so that MS screening adjustments and corrective factors could be applied as appropriate in future pregnancy cases.

Chemokine Receptor Interaction with AFP: Analysis By Computer Modeling

The Chemokine-like Peptide Sites on AFP
Chemokines are chemical attractant cytokines that mediate the migration of cells (i.e. leukocytes, monocytes,macrophages) into and out of tissues and are essential to the immune response and inflammatory reactions [44]. Chemokines are produced locally in tissues and direct the emigration and immigration of cells from the bloodstream into sites of inflammation, infection, and cell proliferation. Chemokines can both direct and influence biological processes such as angiogenesis, degranulation, autoimmunity, HIV infection, tumor growth, parasite infection, and leukocyte trafficking [45]. Discovered less than 20 years ago, the chemokines are now known to function as regulatory molecules in cell maturation, homing, and development in various tissue types. The chemokines comprise a family of nearly 50 ligands and 20 receptors, which despite their size, are remarkably homogeneous with properties similar to IL- 8, the first chemokine discovered. Many chemokines are secreted as a result of pathological conditions while some fulfill regulatory roles; others function in cell migration during histo- and organogenesis and cancer metastasis. Chemokines attracted much attention when it was reported that some receptor family members function as binding sites for the HIV virus.that causes AIDS [46]. Although the chemokine receptors show binding to HIV-1, the main function of chemokines is chemoattraction, ligand binding,cell/tissue homing, and luring cells into tissue parenchyma as found in metastasis [47].
As discussed above, full-length HAFP has been reported to bind to the CCR5 receptor [30,22]. In effect, such binding could cause suppression in growth factor-stimulated cell cancers as well as HIV cell fusion interference suggesting that AFP might mimic certain CCR5 chemokine ligands. AFP has long been implicated in myelopoiesis (bone marrow) and lymphopoiesis during pregnancy and these developmental states are now known to be regulated by the chemokines (Table 2). AFP might be a partial agonist/antagonist chemokine mimic depending on its intravascular and interstitial concentrations and times of exposure. As a mimic, AFP might have an advantage over other chemokine-related competitors in that the others may not be able to act in a decoy ligand fashion. It could be speculated that AFP could serve as a novel type of chemokine , in that AFP could retain its growth regulatory properties but still be able to initiate some of the functions of chemokine ligands (such as immunostimulation). Simply stated, AFP may play a role in physiological or pathological situations in which chemokine-induced actions are involved.
Table 2: The chemokine receptors and ligands are listed together with the receptor family, molecular mass, and length of the amino acid chain. The alphafetoprotein (AFP) amino acid (AA) numbers that interact with the chemokine receptors are also shown.
Previous reports of HIV-1 virus infections in the biomedical literature regarding AFP binding to the CCR5 chemokine receptor have been well-documented and confirmed [30,31]. It follows that if AFP is capable of competing for binding to chemokine co-receptors (i.e., CCR5), it could sterically interfere with virus-cell fusion via interaction with the CD4 + monocyte cell surface membrane during pregnancy. As discussed above and in the Figure 1 legend, many ligand binding and dimerization sites have been localized to the third domain of AFP [48]. By means of protein-to-protein interaction computer software, various amino acid sequences on the AFP third domain can be demonstrated as potential computer “in silico” binding/ interaction sites for the chemokine (C-C) receptor protein family. As demonstrated in Figure 5, the computer-derived chemokine receptor interaction sites on AFP are shown to extend from AA#401 to AA# 571.The receptors that might dock at these plausible AFP amino acid sequence sites include the chemokine receptors CCR5, CCR2, CCR8, CXCR4, and CXCR9. The computer software that generated the AFP binding sites of the chemokine receptors was described in a previous report and was generously provided by Serometrix Biotech LLC, Syracuse, New York [49]. Note that many of the plausible sites of CCR interactions are positioned in the middle of the third domain of AFP with less sites occuring at the N-terminal and C-terminal regions of the third domain. As displayed in Figure 4, the CCR-β and to a lesser extent, the CXCα chemokine ligands reveal cysteine-tocysteine disulfide bridge configurations that resemble those found on the AFP molecule. In fact, it has been reported that AFP shows subdomain molecular mimicry and amino acid identities to certain chemokine ligands from the CC – and CXC - chemokine families [48,50,51]. These data suggest that AFP is capable of interacting with CCR and CXCR receptors due to its molecular resemblance to circulating chemokine ligands of the CCL – and CXCL- families [50]. The published reports [30,31] demonstrating in vitro binding of AFP to the CCR5 chemokine receptor serve to strengthen the present computer docking observations.. The findings that computer software reveals the identity and location of such amino acid CCR5 plausible interaction sites on AFP further expands the scope of knowledge of the prior binding reports. Proteomic subdomain analysis also supports the rationale for such chemokine receptor interaction sites by providing AFP molecular mimicry patterns of the chemokines (Figure 4) which supports the supposition that certain regions of the AFP third domain resemble chemokine (ligand) peptides.
Figure 5: The amino acid sequence from the third domain of human alpha-fetoprotein is displayed in the single letter amino acid code. The NH2 residue is the aminoterminus while the COOH residue represents the carboxyterminus of AFP. The diagram depicts blocks of AFP amino acid sequences that could potentially interact with various chemokine receptors analyzed by protein-to-protein interaction computer software analysis. The diagram demonstrates plausible AFP amino acid interaction sites with various chemokine receptors from AFP AA #401 to 571 with multiple gaps and variable interspacings. The chemokine receptors showing potential sites of AFP interaction include CCR5, CCR2, CCR9, CCR8, and CXCR4, all of which are either obligatory or supplemental co-receptors for the HIV-1 virus. Note that the proposed chemoreceptor interaction sites on AFP roughly correlate with the AFP-chemokine mimic sequences displayed in the Figure and are somewhat clustered between AA #477 to 536. Observe also that the major co-receptors for HIV-1, namely CCR5 and CXCR4 display 7 and 2 interaction sites, respectively, while the supplemental receptors CCR2, CCR8, CCR9 range from 2 to 6 sites. The computer software for protein interactions was supplied by Serometrix Biotech LLC, Syracuse, N.Y. and was described in detail in a previous publication [50]. Dashed-lines indicate overlapping sequence blocks.
It is noteworthy that, in several instances of the CCR- and CXCR- interaction sites on AFP, the receptor interface sites are situated in juxtaposition to a cysteine disulfide bridge configuration (Figure 4). While this appears to be largely true for CCR5 and CXCR4, this cysteine bridge “near neighbor” effect was not always observed for CCR2, CCR8, and CXCR9. One should bear in mind that CCR5 and CXCR4 are obligate co-receptors for the HIV-1 virus binding to CD4-expressing lymphocytes, while other chemokine receptors such as CCR2, CCR8, and CXCR6 are only ancillary or adjunct receptors [52-54]. Chemokine receptors belong to a family of seven-transmembrane spanning G-protein coupled cell surface proteins that engage in intercellular signal transduction. Ligand binding to chemokine G-coupled receptors (GCR) occurs mainly on the excellular loop domains or at the entrance to the base of the transmembrane domain of GCRs. Thus, it is conceivable that the chemokine-like mimicry of AFP in the third domain could provide the underpinning for interaction of the chemokine receptors with AFP as the in vitro reports have demonstrated(see above).

Ataxia Telangiectasia

Ataxia (AT) is an autosomal recessive gene disorder and is one of the chromosome instability syndromes [55,56]. It is characterized by progressive cerebellar ataxia, cutaneous telangiectasia, impaired immunocompetence, increased cell radiosensitivity, and a propensity for the development of lymphorecticular cancers [56,57]. AT further manifests profound immunodeficiency exhibiting sinopulmonary microbial infections, degeneration of the thymus gland, in association with multiple chromosomal aberrations in patients’ immuneassociated cells of the lymphoreticular series [58,59]. AT exhibits chromosomal breakage which encompasses gaps, breaks, dicentrics, and multi-radial configurations. AT is also a neuromotor degenerative disorder which exhibits an incidence ranging from 1 in 40,000 to 100,000 [60,61]. The responsible genetic change has been mapped to one or more mutations on chromosome 11q22-23. Most patients with AT display high AFP serum levels (liver-origin) in most patients [62-64] which can range from 30 to 400 ng/ml [65-67] (Figure 2). Lymphocytes and other cells of AT patients show increased sensitivity to ionizing radiation and to radiomimetic chemicals, displaying aberrant cell cycle checkpoints. These allow continuation through the cell cycle, oblivious to DNA breaks that require repair before the next replication phase can occur [68,69]. This phenomenon is referred to as radio-resistant DNA synthesis (RDS), and is in fact, a basis for one of the phenotypic traits of AT patients [70,71].
Previously, physicians relied mostly on the clinical physical appearances of patients with AT before a diagnosis was made. The use of serum AFP levels and other immune deficiency biomarkers have aided in the diagnostic workup for the early onset of this disorder. Elevated serum AFP is found in over 90% of children with AT (Table 1) and AFP levels are known to increase with age [72,73]. One third of AT patients also display a severe immune deficiency characterized by absent or decreased levels of immunoglobulins including IgA and IgG2 accompanied by the absence of thymic tissue and a decreased responsiveness to skin antigenic stimulation [74-76]. The most serious clinical consequence of patients with AT, however, is their propensity to develop cancer later in life. The lifetime occurrence of cancer in patients with AT is 1 in 20 [77-79]. Lymphoreticular (T-cells) and leukemic cancers predominate in the first two decades, and solid tumors thereafter. Unfortunately, the radiation sensitivity of AT reduces the clinical therapeutic options available to affected patients.
One of the major causes of AT is a gene mutation found in immune system cells such as lymphocytes, monocytes, and fibroblasts; this lesion is located on chromosome location 11q22-23 involving the ataxia telangiectasia mutated (ATM) gene [80,81]. Mutations in this gene have been found in more than 90% of the AT patients examined [82,83]; in addition, more than 400 AT mutations have been documented extending over all 66 exons of this gene [84-87]. Once cloned, the ATM protein was found to be a kinase that shared sequence homology with RAD-3 which regulates passage through the cell cycle after DNA damage occurs and is involved with the high molecular weight phosphoinosital kinase-3 (PI3-kinase) signal transduction pathway [88-90]. The ATM gene also shares homology with various kinases that regulate cell cycle checkpoints following DNA damage, chromosomal abnormalities, increased sensitivity to UV/X-rays, and the presence of chemical mutagens [91] (Table 3). The RAD-3 related PI3 kinase has been cloned and mapped to chromosome 3q22-q24 being named the ataxia telangiectasia RADrelated (ATR) kinase; both ATM and ATR were also found to form part of the synaptonemal complexes produced during meiotic cell division [92]. The ATM kinase activates p53 when cells are subjected to DNA damaging agents such as UV and ionizing radiations; ATM can also phosphorylate c-abl, a protein kinase implicated in the growth arrest response to DNA damage [93,94]. Other ATM protein targets include BRCA1, p95, mdm2, Mre11, and IKB alpha (NF-KB related), all of which are implicated in various stress and apoptotic responses [93-98]. Overall, the ATM kinase appears to be involved in regulating numerous cell cycle checkpoints and apoptotic events in response to damage of double-stranded DNA. The function of the non-mutated normal ATM gene is to sense the presence of doublestranded DNA damage and to mediate an appropriate response. In contrast, the ATM defective gene product is blinded and lacks the ability to correctly process double-strand DNA break repairs which would account for the aberrant cell divisions found in cells of AT animal models, T-cell receptors displaying gene rearrangements in immunoglobulin, somatic recombinations, and in cells responding to DNA damaging agents [99].
Table 3: The DNA-Repair Kinases and Cell Cycle Proteins involved in Ataxia Telangiectasia, Ataxia-Oculomotor Apraxia, and Related Disorders are displayed according to their properties. The AFP amino acid sequences that interact with these kinases “in silico” are displayed in the right hand column.
The neurodegeneration in AT mainly involves loss of a population of Purkinje cells by apoptosis in the cerebellum [100]. However, the neural damage can also occur throughout various brain regions being found in the amygdala, caudate, corpus callosum, thalamic nuclei, and hippocampus 34. However, the neurodegenerative process is not immediately observed in the cerebral cortex, and full-fledged cerebral cell damage in AT does not occur until well after birth with onset being from 2 to 6 years of age [97]. Although AT is a rare disorder, it is found widely distributed through the world in countries such as Turkey, Norway, Costa Rica, Iran, Saudi Arabia, North Africa, South Africa, and the USA [98]. Typically, the total clinical manifestation of AT involves the following characteristics: a) an occurrence of ocular and other telangiectasias, b) growth retardation, c) progressive cerebrallar ataxia, d) neural and immune cell DNA repair disruption, and e) highly elevated serum levels of AFP [99,100]. It is such clinical manifestations that distinguish AT from the other chromosomal instability (mutagen hypersensitivity) syndromes such as Nijmegen Breakage syndrome, Fanconi anenia, Bloom syndrome, xeroderma pigmentosium, and Cockayne syndrome [101,102].
subtype of the AT disorder, referred to as ataxia with oculomotor apraxia (AOA), also exhibits elevated levels of AFP. Apraxia itself is a disorder characterized by loss of learned voluntary motor movements despite having the ability to do so. Apraxia is due to damage in areas of the cerebrum and differs from ataxia which is a lack of coordination of motor movements. Apraxia can affect various areas of the body including the face (mouth, lips) hands and fingers, arms and legs, larynx, and oculomotor muscles. Unlike AT, patients having ataxia with AOA show no increased sensitivity to ionizing radiation; however, similar to AT, AOA is associated with highly elevated serum levels of AFP [103-105] (Figure 2). The gene mutation responsible for AOA resides in the helicase domain of the Senataxin (SETX) molecule [106,107] (Table 3). Thus, AOA is an autosomal recessive inherited disorder with onset over a range of 4 to 14 years in children [104]. The childhood onset of AOA is characterized by progressive neural ataxia, oculomotor apraxia, and the development of motor neuropathy. Cases of AOA have reported worldwide in countries such as Portugal, Japan, French Canada, Italy, France, and the U.S [107].

Interaction of AT and Related Proteins with AFP: Analysis by Computer Modeling

It has been reported that patients with AT exhibit excessively high elevations of serum AFP in both children and adults. Unlike many diseases, AT affects two major body systems, the immune and the nervous system; thus, patients die in late teenage years with few surviving to age 40 [108,109]. As discussed above, the ATM defective gene produces a corrupted kinase which normally regulates passage through S-phase checkpoints of the cell cycle, modulates apoptosis following DNA damage, and promotes signal transduction. Another gene, ATR is a similar-acting kinase[110].Both ATM and ATR protein kinase computer-based “in silico” binding to AFP are discussed below.
The present computer software analysis using protein-to-protein modeling suggests that AFP can potentially interact with both ATM and ATR checkpoint kinases (Table 3). It is not known why AFP serum levels of in AT patients dramatically rise far above those observed in pregnant women (> 500 ng/mL in AT compared to 150 ng/ml in pregnancy, Figure 2). Computer analysis using proteinto- protein interaction software revealed that AFP can potentially bind with both ATM and ATR proteins on the AFP third domain. These plausible amino acid sites of AFP-to-ATM/ATR and other cell cycle protein interactions are depicted in Table 3. AFP amino acid sequences of ATM protein interactions sites begin at AFP AA# 396 and are interspaced at multiple sites ending with AFP# 469 (see Figure 5 for amino acid numbering sequences). At this point, a gap is found which extends to AFP #490. The most intense region of ATM interaction occurs in a molecular loop extends from AA# 489 to 537 even extending to AA# 583. In comparison, ATR interaction sites are less numerous, being concentrated in AFP areas from AA# 480 to 535. As discussed above, the AA# 445 to AA 480 sequences are devoid of AFP-to-ATR AA interaction sites. See Table 3 for other AT-related protein interaction sites on AFP. Such sites of proteinto- protein interactions are presently hypothetical and provide only indirect evidence until they can be proven by in vitro and/ or in vivo experiments. To this end, some reports of AFP-derived peptides from the third domain of AFP may shed some light on the present computer binding results. For example, murine thymic cells treated with AFP-derived peptides overnight, then sumitted to X-ray irradiation showed that such cells were radiosensitized and rendered susceptible to apoptosis as measured by flow cytometry [111]. This study demonstrated that a third domain AFP-derived 34 AA-peptide could induce radiosensitization by prior incubation with and exposure to such peptides. In a second study, these same third domain AFP-derived peptides were administered to MCF-7 human breast cancer cells for 8 days in culture; the lysates from the MCF- 7 cells were then subjected to a global microarray (mRNA) analysis [112]. The microarray analysis revealed that the AFP-derived peptides down-regulated the mRNA in multiple proteins regulating the cell cycle. For example, the checkpoint suppressor-1 (CHES-1) was down-regulated 9.2-fold; cyclin-E was down-regulated by 4.6 fold; SKP2 by 4.3 fold; while DNA replication protein (ESC02) was downregulated 9.2 fold and DNA repair proteins histone-1 and Fanconianemia by 3.2 and 2.0-fold, respectively (Table 3) [112]. Although the microarray findings may not directly confirm the computer software interaction data presented here, they strongly suggest that AFP-derived peptides are involved with the cell cycle and apoptotic proteins and can contribute to down-regulating their RNA message.

Concluding Remarks

It is evident from the above discourse that AFP biomarker levels are elevated in IDDs suggesting an association between AFP synthesis/secretion and deficiencies in the immune system. Although the above discussion included AFP serum levels during pregnancy in women with HIV, IDD patients also display elevated AFP levels in infancy, childhood, adolescence, and adult disorders (Figure 2). The question of why AFP levels are elevated in AT patients is not presently known. AFP in such patients is known to be of hepatic origin [67] and that AFP levels continue to increase with age [72]. Although the AFP serum level increases run parallel to the agerelated worsening of the AT disorder, a link between the degree of neurological disease and AFP levels has not been established. Other than a few clinical oxidative stress and cirrhotic cases, little or no evidence of chronic liver disease/damage has been reported in autopsied AT patients [72] It is of interest that reports indicate p53 and related family proteins act as regulators of AFP gene expression during fetal liver development [73,113]. Thus, the AFP gene in the neonatal AT liver may be under aberrant transcriptional control in conjunction with possible defects in DNA regulatory proteins that are required for hepatic cell maturation. That is, the AFP gene may not be completely turned off by methylation and/or other regulatory control factors. Liver stem cells, which do not secrete AFP, surround the hepatic bile ducts and give rise to the non-parenchymal oval cells capable of secreting AFP. The oval cells further serve as the progenitor cells of the liver parenchymal (non-AFP secreting) cell population. It may be that an abnormal (high) density population of oval cells exist in the childhood/adult AT liver which continue to secrete AFP in the bloodstream.This could be confirmed by histochemical staining and should be pursued. The continued state of immature cells in the liver, verifiable by immmunohistology, might explain the continued elevated levels of AFP in AT patients. Thus, the liver of AT patients may persist as an organ composed of an aberrant, unbalanced mixture of immature and mature cell types composed of stem, oval, and parenchymal cells.
Regarding the AFP protein-to-protein interactions at the molecular level, computer software modeling demonstrated plausible interactions between AFP and chemokine receptors and between AFP and DNA-repair/cell cycle checkpoint proteins. Since the DNA repair proteins are found in lymphoreticular cells bearing T-cell receptors, a link between AT kinases and the immune system may be implied. In the case of the chemokine receptors, computer data are consistent with previously reported in vitro AFP binding to the CCR5 and CXCR4 receptors, which are the co-receptors for HIV cell transmission and infection. However, the interaction of the AFP molecule with DNArepair kinases has not been previously reported in the literature and must remain hypothetical until proven in systems of cell culture or live cell suspensions. In both instances (chemokine receptors and DNA-repair kinases), the “in silico” interactions might be considered as an allosteric interaction since the computer data (not shown) is able to reveal the exact AA sequence of the target protein interaction as determined from the NCBI database protein accession numbers. Tracing such linkages suggests that AFP need not necessarily bind to a major ligand binding pocket, but rather at sites proximal or distal from major docking or binding pockets. As shown in previous reports [112,114,115], the allosteric binding of small peptides or proteins to larger proteins can result in a conformational change in the host protein which can alter ligand binding. These events can be induced as a result of a large protein “jockeying for position” in an attempt to accommodate favorable binding to a ligand, such as a small peptide. This was demonstrated in a prior report involving a time-course kinetic study of estradiol binding to AFP [116]. In this manner, small peptides could serve to act as modulators of ligand docking to major docking sites and/or binding pockets. Thus, AFP peptide binding to allosteric docking sites on proteins (receptors) could not only serve as regulators of binding sites, but could also act to inhibit , reduce, or restrict ligand interactions with such proteins.


The author wishes to extend his thanks to Mr. Andrew Bentley, (Wadsworth Photo & Medical illustration Dept), for his expertise in producing the graphic art illustrations for the Figures of this manuscript


Track Your Manuscript