Altered plasma albumin characteristics and lymphocyte populations in gastrointestinal cancer patients: Detection using modified fluorescence responses

Abstract

The fluorescent probe ABM (an amino derivative of benzanthrone) developed at Riga Technical University (Riga, Latvia) was used to characterize the blood plasma albumin of cancer patients (46 gastrointestinal cancer patients—30 with colorectal cancer in Stages II–IV and 16 with gastric cancer in Stage III) and of healthy controls. The fluorescence intensity of ABM in the blood plasma decreased from the control mean value and showed specific differences in the patients before (i.e., 24 hr pre-) and after (i.e., 10 day post-) they underwent a medically indicated surgical treatment, i.e., a gastric resection or gastroectomy for patients in the gastric cancer group or a colorectal resection for patients in the colorectum cancer group. The significant decrease in the ABM fluorescence in the blood plasma could be explained, in part, by a diminished binding capacity of the albumin of these patients. In fact, before surgery, there was a strong reduction in binding constant (K_a) value for the probe observed in the plasma samples from these patients as compared with the value obtained with the plasma of the healthy donors. The lymphocyte distribution among the subsets also differed between the groups. Surgical treatment affected several immunological parameters and appeared to elevate the functional status of lymphocytes. Interestingly, the ABM fluorescence in the blood plasma was also seen to correlate with select immunological parameters (CD4⁺:CD8⁺ ratios, levels of CD38⁺ cells, etc.) both before and after the patients’ operations. The results in the present study suggest that measures of ABM fluorescence intensity values for plasma albumin and/or especially of lymphocytes (as reflections of their functional activity) might be a useful tool in the evaluation of the immune status of gastrointestinal cancer patients.

Keywords::

• Fluorescent probe
• blood plasma albumin
• immune status

Introduction

Membrane damage in different pathologies may be a consequence not only of lipid peroxidation, but also of protein alterations. Albumin is most abundant in blood plasma and acts as a transport carrier of endogenous and exogenous ligands poorly soluble in water (e.g., certain drugs, fatty acids, etc.) (Grizunov and Dobretcov, 1994; Rolinsky et al., 2007). In general, the principal function of plasma albumin is to transport a wide variety of fatty acids and metabolites via the main binding regions. It is now widely accepted that the dynamics along with structure form the basis of most functional characteristics of biomolecular systems. Albumin also plays a critical role in defining most of the biophysical functional characteristics of plasma and in the maintenance of the immune status of any given organism (Lakowicz, 2000).

It is very important for clinics to receive information on the properties of immune competent cells by an express method. The fluorescent probe proved to be an excellent independent model for such studies (Lakowicz, 1999). This work investigated the possibility of using the fluorescent probe ABM (an amino derivative of benzanthrone) for the detection of structural and functional alterations of blood plasma albumin in pathologies. Such an analysis will have a great potential for determining the control mechanisms associated with the induction and development of the malignant transformed state in the hematopoietic system as well as in the mammalian tissue.

In the present study, we determined the ABM fluorescence intensity in blood plasma from cancer patients in relation to other immunological parameters. By virtue of its sensitivity, selectivity, and large temporal range, fluorescence spectroscopy has become one of the most revealing windows on biomolecular dynamics (Slavik, 1982; Lakowicz, 2000). Our aim was to elaborate a new fluorescent-based method for use in immunologic-based diagnostics of gastrointestinal cancer patients. This type of research is very important in Latvia because, in the context of oncological diseases among the Latvian population in 2006, gastrointestinal cancers ranked in third place in incidence, only surpassed by lung cancers and urogenital tumors.

Materials and methods

Study subjects

The research design used in this study had been approved by the Central Ethics Commission (Riga, Latvia) and all patients enrolled in this investigation had provided their informed consent. For this study, a total of 46 gastrointestinal cancer (i.e., 30 with colorectal cancer in Stages II–IV and 16 with gastric cancer in Stage III) patients (of both sexes) were examined. As a condition for enrollment in these studies, the cancer patients could not have undergone any chemotherapy or radiation treatment(s) before their scheduled surgical procedures (see later). To serve as controls, 14 healthy (average 32–68 years) age-matched volunteers (of both sexes) were employed to ascertain normal (i.e., baseline) levels of lymphocyte fluorescence intensity and other characteristics.

Each of the cancer patients during the course of this study underwent a medically-indicated stomach or colorectal operation to remove all/most of the tumor(s) present. Accordingly, patients in the gastric cancer group underwent a gastric resection or gastroectomy, while those in the colorectum cancer group underwent a colorectal resection. None of the subjects evidenced signs of any other disease(s) or manifested any aberrant post-operative complications (e.g., exaggerated systemic inflammatory responses, fever, etc.) during the course of these studies.

Each patient had their blood sampled 1 day before operation and 10 days after their surgical treatment. In these studies, data from the “whole” gastrointestinal cancer group (i.e., “Common group”; a combination of both gastric cancer and colorectal cancer group patients), as well as the two individual subgroups for each cancer type, were ultimately used in the data analyses reported in this study.

Blood collection

At each patient’s corresponding pre- and post-operative time points, blood was drawn (usually at 8–10 a.m. for fasting values). Then blood was drawn from the healthy volunteers, this was also performed at 8–10 a.m. In all cases, the peripheral venous blood samples were collected into vacutainer tubes containing preservative-free heparin (30 IU per tube); a total volume of ≈7 mL was routinely the target collection volume.

Preparation of peripheral blood lymphocyte suspension and determinations of subsets

Lymphocytes were isolated from the freshly-drawnheparinized venous blood on a Ficol-Verographine gradient, using standard methodologies. After removal of some of the isolated cells for use in the ABM-binding experiments (see later), tagging of the remaining cells with appropriate specific monoclonal antibodies (Becton Dickenson, Stockholm, Sweden) was performed using the company’s protocols. After this labeling, both the absolute numbers and relative percentages of various lymphocytic cell types (i.e., CD3⁺, CD4⁺, CD8⁺, CD16⁺, and CD38⁺) in the samples were evaluated using a Ortho Spectrum III flow cytometer (Ortho Diagnostics Systems, Inc., Westwood, MA). For each blood sample, a total of 25,000 events were ascertained to permit the analyses.

Sample preparation and fluorescence measurements

Investigations in this study, as in earlier studies, were performed using the fluorescent probe ABM (Kalnina and Meirovics, 1999; Kalnina et al., 2006, 2007a,b). This 3-aminobenzanthrone probe (Figure 1) was synthesized by substituting a bromine atom into 3-bromobenzanthrone using an appropriate amine. The chemical structure of the ABM bears a resemblance to the fluorescent probe MBA (3-methoxybenzanthrone), but having a substituted amino group at the 3-position of the benzanthrone molecule. While the use of MBA is limited by two properties [(1) it destroys cells after a short period of time and (2) it fades rapidly in fluorescent light (≈ 80% loss in 6–8 min)], our previous studies have shown that ABM is both photostable and non-toxic to/in cells (Kalnina et al., 2007a).

Figure 1.  Chemical structure of the ABM fluorescent probe. In the structure, X = a morpholino substitution at the site indicated on the ring.

Fluorescent microscopy has revealed that ABM distributes within cell membranes (i.e., plasma, mitochondrial, nuclear), but fails to localize inside the cell nucleus. The environment of the fluorescent probe is quite polar, similar to that of methanol. That addition of Triton X-100 did not lead to any changes in ABM fluorescence intensity, but incubation with cells at 37°C resulted in increased intensity, suggests that there is a localization of the ABM deep into the phospholipid bilayer. Because of these noted utilities (compared to a probe like MBA), we and others have used ABM to characterize peripheral blood mononuclear cells of healthy donors, patients with several non-malignant diseases (advanced lung tuberculosis, multiple sclerosis, rheumatoid arthritis), and those who had been subjected to ionizing radiation during clean-up work at Chernobyl (see Kalnina et al., 2007b for details).

In the present study, the blood plasma (200-fold diluted) or cells (5 × 10⁶ per assay) incubated without probe was used as each patient’s personal “blank” in each experiment. The ABM (resulting concentration in sample = 19.6 μmol/L) was added into a 1 mL aliquot of the patient’s blood plasma (or a 1 mL solution containing the cells) at room temperature (18–20°C) and the mixture then allowed to settle down for ≈5 min. The time interval between cell/plasma isolation and fluorescence measurement was held constant for all samples (i.e., 3 hr). The resulting fluorescence parameters were then registered on a Spectrofluo JY3 spectrofluorometer (ISA JobinYvon Instruments S.A., Longjumeau, France) at an excitation wavelength of 470 nm and an emission wavelength of 520–700 nm. To register luminescence, the sample was placed in a cuvette (1 × 10 × 40 mm³) fixed at an angle of 30° to the excitation light beam. Fluorescence intensity was then recorded and reported in terms of arbitrary units (AU). The final intensity value for each patient’s sample was then generated after accounting for (i.e., subtracting out) the value associated with their personal “blank”; using this approach thereby eliminated any potential contributions from any autofluorescing constituents in the plasma sample.

From the fluorescence data obtained, the value of the ABM binding constant with blood plasma albumin was then calculated. Specifically, the binding constant (K_a) was determined using the graphical method of Klotz (Schulster and Levitzki, 1980).

Statistical analysis

Statistical differences among groups having different spectral characteristics were determined using the Student’s t-test and Mann–Whitney’s U-test. Correlative relationships between spectral characteristics of the ABM and the measured immunologic parameters were determined as outlined by Duncan (1970).

Results

ABM binding with human serum albumin

When the ABM probe was mixed with albumin, no fluorescence shift was observed from the expected zone (i.e., 650 nm), although—as would be expected—a visible increase in fluorescence intensity was seen as compared to that from the buffer solution alone (Kalnina et al., 1996). Changes of pH in the range from 3 to 12 strongly affected the spectrum maximum and fluorescence intensity of albumin-bound ABM (Kalnina et al., 2004). The most prominent changes in fluorescence characteristics occurred at pH values known to cause conformational transitions in proteins (Grizunov and Dobretsov, 1994).

ABM binding with blood plasma albumin

In the gastrointestinal cancer patients, the ABM emission spectra maximum (i.e., at 650 nm)—after combination with the patients’ blood plasma—was not altered in comparison to that seen with the plasma from the healthy control volunteers. In contrast, with respect to fluorescence intensity, before their individual surgical treatments, the average ABM intensity in the patients’ blood plasma was decreased compared to that seen with the samples from the healthy donors. Specifically, the fluorescence associated with samples from the gastrointestinal (“Common”), colorectal, and gastric cancer groups (Figures 2, 3, and 4, respectively) were decreased by 22.5%, 23.0%, and 22.0%, respectively. At 10 days after their operations, the average ABM fluorescence intensity in the samples from the “Common,” colorectal, and gastric cancer groups were decreased further by 13.1%, 13.9%, and 10.3%, respectively. Only in the case of the presurgery value for the “Common” group were the measured average intensity values of the plasma samples from the cancer patients not significantly (P < 0.05) different from the control volunteers’ average value.

Figure 2.  ABM fluorescence intensity in lymphocytes and in plasma from “Common group” patients. Gastrointestinal cancer patients (colorectal cancer, Stage II+III + gastric cancer, Stage III); “Common group” (n = 26). Group number in figure is used to reflect from whom/when samples were isolated [i.e., Group 1: pre-surgery; Group 2: post-surgery; and Group 3: from healthy donors (control group; n = 14)]. Solid bar in each set: ABM fluorescence in lymphocytes; hatched bar in each set: ABM fluorescence in plasma. All intensity values are shown in AU (arbitrary units; mean ± SE). At P < 0.05, *value significantly different from pre-surgical value and/or ^#significantly different from control group value.

Figure 3.  ABM fluorescence intensity in lymphocytes and in plasma from “Colorectal group” patients. Colorectal cancer patients only (Stage II–III) (n = 10). Group number in figure is used to reflect from whom/when samples were isolated [i.e., Group 1: pre-surgery; Group 2: post-surgery; and Group 3: from healthy donors (control group; n = 14)]. Solid bar in each set: ABM fluorescence in lymphocytes; hatched bar in each set: ABM fluorescence in plasma. All intensity values are shown in AU (arbitrary units; mean ± SE). At P < 0.05, *value significantly different from pre-surgical value and/or ^#significantly different from control group value.

Figure 4.  ABM fluorescence intensity in lymphocytes and in plasma from “Gastric group” patients. Gastric cancer patients only (Stage III) (n = 10). Group number in figure is used to reflect from whom/when samples were isolated [i.e., Group 1: pre-surgery; Group 2: post-surgery; and Group 3: from healthy donors (control group; n = 14)]. Solid bar in each set: ABM fluorescence in lymphocytes; hatched bar in each set: ABM fluorescence in plasma. All intensity values are shown in AU (arbitrary units; mean ± SE). At P < 0.05, *value significantly different from pre-surgical value and/or ^#significantly different from control group value.

Whether these observations tracked actual changes in the levels of plasma albumin were also investigated. The results (data not shown) indicate that plasma albumin concentrations (μg/μL) pre-surgery in each of the patient groups were below those in the plasma of the health controls. In each case, these levels (i.e., 70.76 [± 1.07], 71.73 [± 1.34], and 68.00 [± 2.61]—all in μg/μL [mean ± SD]—in the “Common,” colorectal, gastric, and control groups’ samples, respectively) were significantly different from control values (i.e., 83.41 [± 1.16] μg/μL). This meant that the pre-surgery values for albumin only indicated levels in “Common,” colorectal, gastric, and control groups’ samples that were ≈15%, 14%, and 18% below control, while the samples’ fluorescence intensity was correspondingly lower than the control value by 22.5%, 23.0%, and 22.0%.

The plasma albumin concentrations after surgery seemed to be insignificantly impacted (and in the case of the samples from patients in the gastric subgroup, actual rose slightly). Specifically, these values were 68.53 ± 1.36, 68.48 ± 1.78, and 72.29 ± 0.98 (all in μg/μL, mean ± SD) in the “Common,” colorectal, gastric, and control groups’ samples, respectively. It is of interest to note that while the post-surgery albumin values in these three groups’ samples were ≈18%, 18%, and 13% below control levels, their corresponding fluorescence intensities were now even more depressed relative to that of the controls (i.e., by ≈33%, 34%, and 30%, respectively). These results strongly suggest that the noted changes in ABM fluorescence post-surgery in the cancer patients were most likely attributable to some change in the protein(s) themselves rather than due to post-surgery complications (i.e., bleeding or other mechanisms affecting blood volume/composition).

To ascertain whether these results (different ABM spectral characteristics) could be explained, in part, by altered structural characteristics of the plasma albumin in the cancer patients, the average binding constant (K_a) values were determined using the Klotz graphical method. These analyses revealed that the average K_a values for the pre-surgery samples from the patient subgroups decreased strongly from the respective constant associated with the control group (i.e., 1.0 × 10⁵ M⁻¹, colorectal cancer group; 1.3 × 10⁵ M⁻¹, gastric cancer group; 1.8 × 10⁵ M⁻¹, healthy control group). These values represent decrements of from 28% to 45% in the binding of ABM to albumin in the plasma of these patients; the reason for the decrement remains to be fully determined. Interestingly, even though the fluorescence intensity values decreased further for the patients in each subgroup following their surgical procedures, the average K_a values for the samples from the patients increased relative to corresponding preoperative values (i.e., 1.3 × 10⁵ M⁻¹ for the colorectal cancer group and 1.5 × 10⁵ M⁻¹ for the gastric cancer group).

ABM binding with lymphocytes

In the gastrointestinal cancer patients, the ABM emission spectra maximum (i.e., at 650 nm) after combination with the patients’ lymphocytes (as with their plasma) was not altered in comparison to that seen with the cells from the healthy control volunteers. With respect to fluorescence intensity, before their individual surgical treatments, the average ABM intensity in these cancer patients’ blood plasma was decreased as compared to that seen with the samples from the healthy donors. Specifically, the fluorescence values associated with the samples from the “Common” and gastric cancer groups (Figures 2 and 4, respectively) were decreased by 16.0% and 22.0%, respectively. This value for the gastric patients was significantly (P < 0.05) different from the control. In somewhat of a surprise, the average value noted with the cells from patients in the colorectal subgroup was actually 12.0% greater than the control level (Figure 3); however, even with this increase, the value was not significantly different from the average control value.

In contrast to what was observed with the plasma samples, the average ABM fluorescence intensity values noted with the cells from patients at 10 days after their operations were greater than the values seen with the control volunteers’ cells. Specifically, the values associated with the cell samples from the “Common,” colorectal, and gastric cancer groups were greater than control levels by 24.0%, 44.0%, and 16.0%, respectively; only the gastric subgroup value failed to reach statistical significance (P < 0.05). In comparison to the pre-operative values, these average ABM fluorescence intensity values at 10 days after the patients’ operations had increased by 47.6%, 28.6%, and 81.2%, respectively, in the samples from patients in the “Common,” colorectal, and gastric cancer groups.

Lymphocyte count and subpopulations

The absolute number of CD3⁺ cells and CD16⁺ natural killer cells; the relative percentages of all lymphocytes, CD16⁺ and CD38⁺ cells; as well as the CD4⁺:CD8⁺ ratio in the blood samples of the healthy volunteers and of the cancer patients, before and after each underwent their operations, were determined. The results show that among the patients in the “Common” group (Table 1), before surgery, only the levels of CD3⁺ cells, the relative percentage of CD38⁺ cells, and the CD4⁺:CD8⁺ ratio were significantly decreased (i.e., by 15.7%, 79.5%, and 27.7%, respectively) as compared to corresponding values associated with the control subjects. Within this group, after surgery, these same three parameters were still significantly lower than the corresponding control values. However, the number of CD3⁺ and CD16⁺ cells, as well as relative percentages of all lymphocytes and CD16⁺ cells, were significantly reduced (i.e., by a further 4.7%, 52.9%, 24.1%, and 42.4%, respectively) relative to pre-surgery levels. Interestingly, the relative percentage of CD38⁺ cells and the CD4⁺:CD8⁺ ratio increased compared to the patients’ pre-surgical values, but these did not reach levels of the controls (i.e., still were 56.6% and 17.6% lower than control values, respectively).

Table 1.  Peripheral blood lymphocyte subpopulation counts in gastrointestinal cancer patients (i.e., “Common group”).

Gastrointestinal cancer patients (colorectal cancer, Stage II+III + gastric cancer, Stage III); “Common group” (n=26)
	Lymphocytes (%)	^aCD3^+	CD16^+%	^aCD16^+	CD38^+%	CD4^+:CD8^+
^b1	26.80 ± 2.01	^‡1116.21 ± 119.20	13.48 ± 3.01	280.41 ± 30.12	^‡5.04 ± 0.97	^‡1.36 ± 0.07
2	*^‡20.33 ± 1.47	*^‡1063.77 ± 113.62	*7.76 ± 1.09	*132.14 ± 21.08	*^‡10.68 ± 2.06	*^‡1.55 ± 0.10
3 (Controls)	28.00 ± 1.30	1324.79 ± 115.41	12.50 ± 1.10	389.00 ± 24.11	24.60 ± 1.60	1.88 ± 0.16

^aValues shown are in terms of absolute numbers (mean ± SE).

^bIndicates when or from whom samples were isolated: (1) before surgical treatment; (2) after surgical treatment; and (3) healthy donors (control group; n = 14).

*Value significantly different from pre-surgical value (P < 0.05); ^‡significantly different from control group value (P < 0.05).

The results among the patients in the “colorectal” group (Table 2) indicate that, before surgery, the numbers of CD16⁺ cells, the relative percentage of CD38⁺ cells, and the CD4⁺:CD8⁺ ratio were each significantly decreased (i.e., by 19.2%, 86.2%, and 32.4%, respectively) as compared to corresponding control subject values. Somewhat unexpectedly, relative percentages of CD16⁺ cells in this group were actually significantly greater (by 27.6%)—and the relative percentage of lymphocytes no different—than in the blood of the control volunteers. Within this same group, after surgery, the number and percentages of CD16⁺ cells, as well as the percentages of all lymphocytes, were each significantly reduced (i.e., by 53.9%, 50.3%, and 16.1%, respectively) relative to corresponding pre-surgery levels. Again, the relative percentage of CD38⁺ cells and the CD4⁺:CD8⁺ ratio increased (by 27.4% and 18.1%, respectively) compared to pre-surgical values, but again did not reach control levels (i.e., still were 48.4 and 20.2% lower, respectively).

Table 2.  Peripheral blood lymphocyte subpopulation counts in the distinct subsets of the study’s cancer patients.

Colorectal cancer patients only (Stage II–III) (n=10)
	CD16^+%	^aCD16^+	CD4^+:CD8^+	CD38^+%	Lymphocytes (%)
^b1	^‡15.95 ± 2.18	^‡314.18 ± 39.27	^‡1.27 ± 0.10	^‡3.40 ± 1.20	28.00 ± 1.30
2	*^‡7.93 ± 1.43	*^‡144.75 ± 22.44	*^‡1.50 ± 0.13	*^‡12.70 ± 3.40	*^‡23.50 ± 2.20
3 (Controls)	12.50 ± 1.10	389.00 ± 24.11	1.88 ± 0.16	24.60 ± 1.60	28.00 ± 1.3
Gastric cancer patients only (Stage III) (n=10)
^b1			^‡1.27 ± 0.10	^‡15.30 ± 0.80	^‡32.80 ± 2.85
2			*^‡1.55 ± 0.13	*^‡17.60 ± 1.60	*^‡21.96 ± 1.60
3 (Controls)			1.88 ± 0.16	24.60 ± 1.60	28.00 ± 1.30

^aValues shown are in terms of absolute numbers (mean ± SE).

^bIndicates when or from whom samples were isolated: (1) before surgical treatment; (2) after surgical treatment; and (3) healthy donors (control group; n = 14).

*Value significantly different from pre-surgical value (P < 0.05); ^‡significantly different from control group value (P < 0.05).

Finally, the results from the patients in the “gastric” group (Table 2) show that, before surgery, the relative percentage of CD38⁺ cells and lymphocytes were each significantly decreased (i.e., by 32.4% and 37.8%, respectively) compared to control values. Interestingly, the relative percentage of lymphocytes was significantly greater (by 17.1%) than in the blood of the control volunteers. After surgery, the percentage of lymphocytes was significantly reduced (i.e., by 33.0%) relative to the pre-surgery levels; this shift was sufficient enough to yield a value that was now significantly lower than control levels (i.e., by 21.6%). As with the results in the other patient subgroups, the relative percentage of CD38⁺ cells and CD4⁺:CD8⁺ ratio increased (by 15.0% and 18.1%, respectively) compared to pre-surgical values, but did not reach control levels (i.e., still were 28.5% and 17.6% lower, respectively).

Examination of relationship between ABM fluorescence and blood lymphocyte profiles

Pre-operation, with each of the patient groups, the ABM fluorescence intensity was found to correlate with the relative number of CD38⁺ cells (r² = +0.958, +0.956, and +0.859, respectively). After surgical treatment, in the “Common” (i.e., gastrointestinal + colorectal cancer) group, a reverse correlation of ABM fluorescence intensity was found with respect to the relative count of CD16⁺ cells (r² = −0.791 and r² = −0.498, respectively); however, in the gastric cancer group, it was held with respect to the relative count of CD8⁺ cells (r² = −0.804) (Tables 1 and 2). After the operations, both the CD4⁺:CD8⁺ ratio and relative number of CD38⁺ cells in gastrointestinal patients’ blood was observed to be increased.

Discussion

The novel fluorescent probe ABM (an amino derivative of benzanthrone) localizes deep within the phospholipids bilayer of lymphocytes membrane. Thus, in studies with lymphocytes, it can be concluded that changes in the spectral parameters of ABM (i.e., shifts in magnitude of fluorescence or actual wavelength associated with normal maximal fluorescence [i.e., F_max]) could reflect modifications in one/more interdependent (i.e., inter-related) properties of the cells. These could include the lymphocytes’ (1) outer membrane physicochemical state, (2) membrane microviscosity, (3) proliferative activity, (4) lipid metabolism, and/or (5) phenotypical profile. As seen in the studies mentioned here, while the noted changes in the studied parameters (i.e., fluorescence behavior) could be useful in reflecting alterations in lymphocytes of the cancer patients in each subgroup (at both pre- and post-surgical stages), they may also ultimately be of use as potential indicators of alteration in cellular immunity in these individuals. Follow-up studies are underway to see whether this concept can be validated.

We also sought to ascertain whether shifts in ABM binding with plasma albumin could potentially be utilized as a part of an overall preliminary immunodiagnostic screening test in cancer patients. The choice to examine albumin, among the myriad of constituents in plasma, is that this protein is practically the single source of ABM binding and subsequent fluorescence in plasma (Grizunov and Dobretsov, 1994). Our earlier studies showed that within plasma, albumin is nearly alone in binding with ABM with a very high level of selectivity (Kalnina et al., 1996, 2004, 2007a). The distribution of ABM fluorescence (intensity) within fractions of human plasma was seen to be albumin >>> globulins >> non-specific binding by other components (i.e., 90%, ≈ 5%, < 1%, respectively). These widely disparate binding results were confirmed in studies wherein exogenous globulin was added to plasma samples and there was no shift in fluorescence intensity or F_max. Clearly, only significant shifts in albumin levels OR alterations/conformational changes in albumin itself seemed to have a major impact on these ABM fluorescence endpoints.

In the present study, the differences in total albumin concentrations, pre- and post-surgery, among the cancer patients in each group did not seem to correlate well with the relative changes in ABM fluorescence (relative to values in control subjects’ plasma). This apparent “extra diminution” in fluorescence strongly suggested that there was either a novel competition for probe by other substances in the patients’ plasma or that the albumin in these patients had undergone modification(s) that affected its ability to bind ABM. The fact there were substantive changes in binding constant (K_a) values lends support to the latter viewpoint. However, this finding in and of itself does not outright preclude the possibility of the former event having occurred as well.

These shifts in ABM binding constants in the plasma samples from the cancer patients, as noted earlier, could be due to a generic decreased binding by/conformational changes in their albumin molecules. Structural or functional alterations of albumin could be manifest as “shifts” away from normal “main” binding sites with high affinity for the probe to other binding sites with far lower affinities and specificities. Such shifts would be in agreement with the observations of Togashi and Ryder (2006, 2007) that albumin molecules are known to contain different binding sites (i.e., classes) for various probes. As Petitpas et al. (2001b, 2003) noted, albumin normally carries a variety of endogenous ligands like nonesterified fatty acids, bilirubin, and thyroxine; however, this protein can also bind an impressive array of drug molecules, including warfarin, ibuprofen, and indomethacin, as well as their metabolites (Petitpas et al., 2001a). It seems very likely that patients in the groups in the present study had ingested painkillers (both prescribed and retail) during the course of their disease; thus, a presence of these drugs/metabolites on their albumin could have contributed to the noted shifts in ABM fluorescence/K_a values. Our future studies will endeavor to recruit non-cancer patients with a “similar” history of painkiller intake in order to ascertain whether this was a main reason underlying our observations (regarding the albumin outcomes) or if there is something more inherently unique to the patient’s cancer-bearing status that influenced the measured endpoints.

This second standpoint is not without foundation. In oncopathology, the blood plasma content of two important unsaturated fatty acids (i.e., oleic acid and arachidonic acid) is increased, and these natural constituents also increasingly occupy binding sites on albumin (Grizunov and Dobretsov, 1994). Both are observed to occupy binding sites distributed across the protein that happen to also be bound by medium or long-chain saturated fatty acids. The resulting restrictions imparted on the binding configurations of the protein would then account for shifts in the binding affinities at the primary sites between polyunsaturated fatty acids and their saturated or mono-unsaturated counterparts (Petitpas et al., 2001). It remains to be determined whether these alterations in fatty acid composition/binding also result in conformational changes in the albumin that impact upon ABM binding to its major (high selectivity) binding sites.

As noted earlier, changes in fluorescence parameters of the cancer patients’ lymphocytes could be reflective of changes in one/more inherent characteristics of their cells. In these studies, at least two, that is, proliferative activity and phenotypical character, could readily, albeit indirectly, be evaluated by examining changes in lymphocyte populations (i.e., their numbers) themselves. While the flow cytometry studies did indicate significant changes in lymphocyte (and subpopulation) levels among the cancer patients, unfortunately, the studies failed to yield overall lymphocyte (or subtype) population patterns that paralleled the concurrent changes in ABM fluorescence (i.e., Table 1 vs. Figure 1, example of this “lack of comparativeness”). Among all the subpopulation endpoints reported, only those of “CD38⁺%” and the “CD4⁺:CD8⁺ ratios” approached reflecting trends seen with the patients’ fluorescence measurements. Specifically, the pre-surgery levels of each of these cytometric values were “maximally” reduced relative to the control subjects’ values; post-surgery, these two values were increased, but in contrast to the fluorescence levels, these values did not reattain (or surpass) counterpart control levels. In light of the cancer patients’ post-surgical (1) persistent lower numbers of lymphocytes (both total and within subclasses) and (2) fluorescence values that were uniformly significantly greater than in control subjects’ cells, we surmise some factor(s) about these patients’ lymphocytes (i.e., some undefined phenotypical characteristics) can cause amplification of the ABM fluorescent response.

The fact that this “disconnect” between these two parameters is most predominant during the post-operative period strongly suggests that these as yet-undefined modifying factors in the cancer patients might be related to their general immune response to the surgical procedure. Our future studies will need to recruit non-cancer patients with a “similar” history of surgical intervention/protocols (such as among patients suffering enterocolitis, undergoing local biopsies for non-cancer disorders, etc.) to ascertain whether the surgical procedure itself was a main reason for our observations (regarding the “disconnect”) or whether, as with the albumin findings, there is something more inherently unique to a cancer-bearing status that influenced the measured endpoints.

As expected, the CD4⁺:CD8⁺ ratios were seen to be increased in the cancer patients after they had undergone their respective operation. This would be expected as it is well accepted that CD4⁺ helper cells stimulate and CD8⁺ (suppressor and cytotoxic) cells inhibit the immune response during the healing process. While that explanation for any potential changes in the phenotypic characteristics of these patients’ lymphocytes is somewhat straightforward, what is less clear is the basis for the post-surgical increase in CD38⁺% values and why, to begin with, they are lower than in the control groups. This is because, most often, increased levels of CD38⁺ cells are associated with patients suffering with lymphocytic leukemias than with the solid tumors (such as those associated with gastric/gastrointestinal cancers; Palmer et al., 2008).

In general, CD38 is expressed primarily on B-lymphocytes and T-lymphocytes, as well as stem/germ cells (Cesano et al., 1998; Lund et al., 1999); the CD38 ligand is an ADP-ribosyl cyclase enzyme that regulates the activation and growth of these lymphoid (as well as myeloid) cells (reviewed in Deterre et al., 2000; Wykes et al., 2004; Sandoval-Montes and Santos-Argumedo, 2005). The data in the current study clearly show no evidence of any B-lymphocyte-based leukemia (i.e., CD16⁺ cell levels were lower in patients’ pre- and post-surgery blood samples than in controls) among the cancer patients. Thus, we conclude that the increase in CD38⁺ cell levels is more probably due to an increased presence of CD38⁺ T-lymphocytes. While Funaro and colleagues almost 20 years ago first described the role of CD38 in stimulating the activation and proliferation of these cells (e.g., Alessio et al., 1990; Funaro et al., 1990), for the above-noted reasons, we conclude from our findings that the increase in CD38⁺ cell levels post-surgery was not likely due to absolute increases in T-lymphocytes, but in their activities. Such an outcome would be in keeping with the changes in the fluorescence values for these lymphocytes. For this premise to be valid, apart from showing that there are increases in relative levels of CD38-bearing T-lymphocytes due to activation during the post-surgery healing process, there still needs to be an explanation as to why these cells’ levels were initially lower in the patients than in the controls.

One potential explanation is in the biology of the tumors themselves, that is, they are solid tumors of the gastrointestinal system that impact on a wide variety of local cell types, including the endothelium. This particular cell type in the gut is of interest here in that there appears to be a critical relationship among endothelial cells, CD38 expression, and activation of T-lymphocytes (i.e., CD4⁺CD45RA⁺ cells; see Dianzani et al., 1994; Deaglio et al., 1996). It is plausible that normal interactions between T-lymphocytes and endothelium (Dianzani and Malavasi, 1995) are likely “interrupted” simply as a result of changes in accessibility (secondary to alterations in gut architecture as tumor grew). A lack of lymphocyte–endothelium interactions could help explain why there was a diminution in CD38⁺ cell levels before surgery; during the post-surgery recovery, angiogenic processes (i.e., during microvasculature repair/reformation at wound site) would allow for an increase in these particular cell–cell interactions—in particular, with a population of endothelial cells in very active states during the reparative processes. Future histopathology studies using biopsied samples from the gastrointestinal tracts of patients with cancers and those that underwent biopsies for non-cancer-based reasons (see earlier comments) should be useful in allowing us to verify the degree of these hypothesized cell–cell interactions.

Apart from potential changes in lymphocyte–endothelium interactions as contributing factors for the reductions (vs. controls) in CD38⁺ cell levels—and their “recoveries” after surgical removal of the tumor—in the cancer patients, there are other possible reasons for these two observations. Among these, specifically, is the fact that patients with colorectal/gastrointestinal cancers (especially those at more advanced stages) tend to have significant levels of circulating interleukin (IL)-4 (Sharma et al., 2008). This is critical in that it has been shown, at least with B-lymphocytes, that exposure of these cells to IL-4 reduced the amount of CD38 antigen on and in these cells (Shubinsky and Schlesinger, 1996); no evidence was obtained for accelerated breakdown, shedding, or internalization of CD38 molecules, or for the accumulation of CD38 molecules in the cell interior, due to IL-4. In our ongoing studies, we will analyze patients’ blood samples for IL-4 both pre- and post-surgery to see whether its levels reflect the observed changes in the CD38⁺ lymphocytes and their fluorescence responses (indicative of phenotypic changes likely related to activation) in the presence of ABM.

The results of the ABM studies presented here show that, as might be expected, the presence of solid tumors and surgical interventions can affect the functional activity of lymphocytes. These results are in agreement with previously-performed investigations to characterize the outer cell membrane of lymphocytes of cancer patients, patients with autoimmune disease (i.e., rheumatoid arthritis), and workers who had been contaminated during the clean up at Chernobyl (Kalnina et al., 2004, 2005, 2007b). Likewise, the observed changes in the ABM spectral parameters in blood plasma are probably coupled with alterations in cellular mechanisms of immune regulation in the patients here. Taken together, all the results in the present study showed that measures of ABM fluorescence intensity values for plasma albumin and/or especially those of lymphocytes (total and among different subtypes) could potentially be a useful tool in clinical immunological screenings to estimate the immune status of gastrointestinal cancer patients. Compared to many other commonly-used diagnostic protocols, this fluorescence-based method is less expensive and not very time consuming, technically simple, and 100 times more sensitive than standard absorbance-based methods. Ultimately, a final determination of whether these types of measures extend to other human disease states remains to be determined. Ongoing studies are seeking to answer this very question.

Acknowledgments

This work was supported by the Latvian Council of Science, Grant Number 09.1209.

Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.