Detection and follow-up of a soluble alpha-haemoglobin pool in the red cells of stored blood units

INTRODUCTION

One of the key objectives of red blood cell (RBC) transfusion is to increase the haemoglobin (Hb) concentration to improve oxygen delivery to the tissues. During their 120-day lifespan in the bloodstream, red blood cells (RBCs) undergo morphological and physiochemical modifications^1,2 that are also observed in blood units as “storage lesions”. Alterations include the rigidity of the RBC membrane, the clustering of the anion exchanger Band3, the appearance of neoantigenic domains targeted by senescence autoantibodies, and the externalisation of anionic phospholipids³. Preservative solutions may slow these changes in blood units that are stored for up to 42 days in banks^4,5 to conserve the optimal deformability of the RBCs in blood units. However, they progressively accumulate these irreversible “storage lesions”, thus altering their half-life after transfusion^6–8. RBCs are highly susceptible to oxidative stress and free radical-induced alterations⁹, and Hb is the molecule most affected by this stress. Human adult Hb (HbA α2β2) has both oxygenated and deoxygenated forms, each with its own characteristic absorbance spectrum. Only α2β2 tetramer can efficiently deliver oxygen, since the β-Hb and α-Hb individual subunits have too strong an affinity for oxygen to release it.

The presence of a soluble α-Hb pool has previously been shown in fresh RBCs drawn from healthy subjects with the specific quantitative method we developed¹⁰. However, the detection of such isolated Hb subunits and the changes they undergo has never been explored in RBCs stored in blood units under refrigerated storage over time. Here we investigated the presence of soluble α-Hb pool in RBC units (also named “blood units”) at different times during the storage period, and the impact of ageing of the blood unit on the amount of its α-Hb pool. We then tested the effect of a cryopreservation period on the α-Hb pool, a process commonly used in blood banks to preserve RBC units with rare blood group phenotypes.

MATERIALS AND METHODS

This research was performed in accordance with the Declaration of Helsinki and was approved by our institutional Ethics Committee (CPP no.11-047).

Processing and conservation of blood units, preparation of RBC lysates

RBC units from whole blood were processed at the Etablissement Français du Sang (EFS) Preparation Unit according to the European guidelines¹¹. The twenty-one units selected arrived 2–3 days after blood collection and were stored at temperatures from +4 to +6°C in a standard blood bank refrigerator. The EFS verified that none of the selected units had the sickle-cell trait but a deletion in one or two of the four α-globin genes cannot be totally excluded. Samples were removed 3–8 days (day [D]3–D8) and 38–42 days after collection (D38–D42); for sixteen RBC units, additional samples were collected at two other intermediate time points (D13–D17 and D24–D29). Samples of four RBC units at D3–D8 were tested by a cryopreservation/ thawing process using a glycerol solution, as previously described¹². For the preparation of RBC lysates, aliquots from the selected blood units were centrifuged at 2,880 g for 10 min (Figure 1B) and the RBC fractions frozen until use. RBCs were thawed and lysed with cold distilled water. The mixture was incubated on ice, centrifuged at 16,000 g at +4 °C, and the lysates recovered in the supernatant.

Figure 1.

Overview of the alpha-haemoglobin (α-Hb) pool measurement

(A) The assay used recombinant α-haemoglobin stabilizing protein (AHSP) to capture the soluble α-Hb present in the red blood cell (RBC) lysates. Recombinant AHSP was produced as a GST-AHSP fusion protein in E. coli BL21(DE3) and purified by affinity chromatography. (B) RBCs were removed from leukoreduced RBC units at the beginning (day [D]3–D8) and end (D38–D42) of storage at temperatures from +4 to +6 °C. Four RBC units drawn at D3–D8 were tested for cryopreservation by comparing samples at D3–D8 before and after freezing for 15 days at −80 °C. RBC fractions were separated by centrifugation and stored at −80 °C. The RBCs were thawed and lysed with water, incubated for 30 min on ice before centrifugation at 16,000 g for 30 min and RBC lysates recovered in the supernatant. (C) GST-AHSP was bound to a glutathione-Sepharose 4B-coated 96-well filter plate. The wells were then incubated with RBC lysates, to capture the soluble α-Hb. Finally, the bound proteins were eluted with reduced glutathione. After elution of the GST-AHSP/α-Hb complex, the α-Hb pool was quantified by measuring absorbance at 414 nm and expressed in ppm, equivalent to ng α-Hb/mg of total Hb subunits/mL RBC lysate. The experiment was performed twice for each sample.

Preparation of GST-AHSP and detection of the α-Hb pool

The α-Hb dosing assay uses the specific character of the interaction between the α-Hb and the α-haemoglobin stabilizing protein (AHSP), the α chaperone, to trap the soluble α-Hb in RBC lysates. The recombinant AHSP was produced with glutathione S-transferase (GST-AHSP) in E. coli and purified by affinity chromatography (Figure 1A), as previously described¹³. Detection and measurements of the soluble α-Hb pool was also performed as previously described¹⁴. Briefly, the RBC lysates were applied to filter plates coated with the GST-AHSP coupled to glutathione sepharose (Figure 1C). Plates were washed and the bound proteins eluted with reduced glutathione in Tris buffer. The α-Hb in the eluted fraction and total Hb in haemolysates (expressed on a haem basis) were determined on a spectrophotometer in the Soret band at 414 nm. The α-Hb pool was expressed in ppm¹⁰ equivalent to ng α-Hb/mg of total Hb subunits/mL lysate.

Statistical analysis

Quantitative variables are expressed as the arithmetic mean±standard deviation (SD). Data were analysed with the Prism 6.0 software (GraphPad Inc., San Diego, CA, USA). p<0.05 was considered statistically significant.

RESULTS

The presence of a soluble α-Hb pool was investigated in the lysates from the selected blood units (n=21). An α-Hb pool was detected from the beginning of storage (D3–D8), just a few days after the preparation of units from whole blood (Figure 2A) with a mean value of 126±23 ppm; α-Hb pool values ranged from 72 to 165 ppm, with a dispersion from the different blood units tested (interquartile range 33). At the end of storage (D38–D42), the Hb pool values were significantly higher (p<0.0001) with a mean value of 152±29 ppm and a higher dispersion (114–209 ppm; interquartile range 48.5). For further analysis, the values for the soluble α-Hb pool were also obtained at two additional time points midway through the storage period of the same RBC units (n=16); at D13–D17 and D24–D29, quantified mean amounts were 131±30 ppm and 134±34 ppm, respectively (Figure 2B). For four RBC units, the effect of a short period of freezing on α-Hb values was also assessed by comparing samples selected at D3–D8 before (97±24 ppm) and after (129±10 ppm) freezing for 15 days at −80 °C. α-Hb pool values of these samples were only slightly higher (p=0.25) (Figure 2C).

Figure 2.

Detection and follow-up of the soluble alpha-haemoglobin (α-Hb) pool in red blood cell (RBC) lysates

(A) Detection of an α-Hb pool in RBC lysates. Leukoreduced RBC units (n=21) stored at temperatures from +4 to +6 °C were measured at the beginning and end of storage (day [D]3–D8 vs D38–D42); each α-Hb value is the mean of two measurements. Statistical analyses were performed with Wilcoxon matched-pairs signed-rank test. Results are shown as box-and-whisker plots with individual values indicated as dots; horizontal bars indicate the median. (B) Effect of the storage time: changes in the α-Hb pool values were measured at four different time points in lysates from RBC units (n=16), at D3–D8, D13–D17, D24–D29 and D38–D42. Each symbol represents a different RBC unit; each α-Hb value is the mean of two measurements. Statistical analyses were performed with Friedman’s test followed by Dunn’s multiple comparisons test. (C) Effect of freezing/thawing on the α-Hb pool: measurements were made on four lysates from freshly prepared RBC units at D3–D8 and from the same units after freezing at −80 °C and thawing 15 days later. Each symbol represents a different RBC unit; each α-Hb value is the mean of two measurements. *p<0.05, ***p<0.001, ****p<0.0001.

DISCUSSION

Previous studies by our team¹² on the same blood units stored for 42 days showed no significant changes in RBC volume, osmotic resistance, or mean corpuscular Hb concentration (MCHC) over time, providing evidence that conventional storage of RBC units did not modify RBC rheology. By contrast, and as expected, the pH of the unit supernatants decreased rapidly^3,12.

The presence of isolated Hb subunits during storage of RBCs units has never previously been studied. An α-Hb pool is detected in blood units from the beginning of storage (D3–D8) at temperatures from +4°C to +6 °C, increasing over the 42-day storage period. The term “soluble α-Hb pool” corresponds to the α-Hb not bound to β-Hb that can be linked to AHSP in RBCs¹⁰. The detection of such an α-Hb pool in blood units is not surprising, given that the presence of α-Hb has already been reported in RBC lysates obtained from healthy volunteers with a normal Hb phenotype (n=38)¹⁴; in that context, the mean value was 81±15 ppm, with a lesser degree of dispersion (54–115 ppm; interquartile range 21).

The differences observed in α-Hb pool values between that of RBC units and that of freshly prepared RBCs from healthy volunteers can be explained by the difference in storage temperatures after collection, in the use of different anticoagulants, or in the sample preparation. In fact, for the preparation of RBCunits, whole peripheral blood is collected into citrate phosphate dextrose (CPD)-anticoagulated bags and kept at room temperature for 2–20 hours before processing¹¹. Leukoreduction is then performed by filtration before the transfer of the RBCs to a bag containing saline adenine-glucose-mannitol (SAGM); all these steps are performed at room temperature, taking a mean time of 10–24 hours, before storage at temperatures from +4°C to +6°C in blood banks. By contrast, the preparation of fresh RBCs drawn from volunteers was processed at +4°C within two hours of collection on ethylenediaminetetraacetic acid (EDTA) and no leukoreduction step was carried out¹⁰. Furthermore, we have observed that α-Hb pool values tend to increase when the whole blood sample is stored for a period of time at room temperature (unpublished observations). It has been also reported that temperature influences the kinetics of dissociation of the αβ dimers into α and β monomers: an increase in temperature from +7°C to +37°C resulted in a 50-fold increase in the dissociation rate constant¹⁵. All these data indicate that an increase in temperature, along with the duration of the procedure in processing the RBC unit, can have an impact on the α-Hb pool values detected and may explain the increase in the α-Hb pool in blood units.

An excess of α-Hb chains that, by precipitating on the RBC membrane and acting as active oxidants, leads to oxidant damage has previously been reported^16,17. In our team, we initially detected the α-Hb pool in RBCs drawn from pathological β-thalassemia blood samples^10,14. In this Hb disorder, an imbalance in the biosynthesis of globin chains lead to an excess of α-chains. Precipitation of α-chains, and oxidative damage in erythroid precursors and RBCs, resulting in inefficient erythropoiesis, have all been observed. In the most severe forms of β-thalassemia, the α-Hb pool values are higher than 1,000 ppm (very high in comparison to the α-Hb pool observed in RBC units) and this correlates well with the clinical severity of the disease. Here, the detected α-Hb pool value is negligible compared to the amount of functional Hb in RBC units and would have had almost no impact on the quality of the stored RBC units. Furthermore, most blood units are transfused to patients between D8 and D18, and the α-Hb pool remains practically stable within this timeframe (Figure 2C), thus supporting the view that the values of α-Hb would not affect the choice of RBC units to be transfused.

The wide dispersion of values from different RBC units observed throughout the storage period, but also increasing towards the end (interquartile range 33 at D3–D8 vs 48.5 at D38–D42), may reflect the well-known variability between different blood bags obtained from the same donor⁸. The units were verified for the lack of the sickle-cell trait but were not genotyped for globin genes. In a previous study, out of 50 healthy volunteers genotyped for globin genes, 20% had an abnormal α-globin genotype and α-Hb pool values lower than those observed in normal α-globin subjects¹⁸. This could be due to the presence of an α-thalassemia mutation¹⁰. Thus, it would be of particular interest to know more about the α-globin genotype of those two RBC units with α-Hb values lower than the average of the other units tested (Figure 2B; see at D3–D8).

It is important to remember that the use of cryopreserved blood units with rare phenotypes can be required for transfusion in certain circumstances, particularly for sickle-cell anaemia (SCA) patients in painful acute crisis or those experiencing severe haemolytic episodes¹². The results we obtained before and after freezing some blood units clearly showed that the α-Hb pool increased only slightly. Thus, such a freezing/thawing procedure required to transfuse previously cryopreserved rare blood units does not significantly modify the soluble α-Hb pool. Finally, as a control, we verified the impact of storage time on Hb functionality in stored RBCs (Online Supplementary Figure S1) with spectra evaluated for soluble Hb in cytosols of RBCs from blood units. The characteristics of soluble Hb were correctly preserved during a lengthy period of refrigerated storage, as shown for other RBC parameters in previous rheological studies by our team^8,12.

CONCLUSIONS

We show for the first time the presence of a soluble α-Hb pool in blood units and its significant increase over a 42-day refrigerated storage period. This work, monitoring the amount of soluble α-Hb in blood units, tested here on a total of twenty-one units, now require confirmation on a larger study sample. Whether the increase in the α-Hb pool due to storage may affect the quality of blood to transfuse remains an open question. An excess of α-Hb chains, by precipitating on the inner side of the RBC membranes, operating as active oxidants, may lead to oxidant damage^16,17; however, the final quantity of the α-Hb pool we detected remained low and this would have a negligible impact in terms of the RBCs ability to fulfil their oxygen carrying function.

It would be of interest to investigate whether the α-Hb pool could be a supplementary quantitative parameter to be used to assess the quality of blood units for transfusion. If this proves to be the case, this biomarker should be further tested as its increase over a determined cut-off threshold value could be a useful parameter in transfusion medicine in order to improve the selection of RBC units in blood banks. A threshold above 150 ppm¹⁰ in stored RBCs might be considered unsuitable to transfuse acute patients, such as SCA patients experiencing haemolytic events or patients with severe sepsis. Such a new quality parameter would help in the selection process of blood units in banks for transfusion recipients who are highly dependent on the quality of blood products they receive. Although an attractive hypothesis, the feasibility of this does, nevertheless, remain to be demonstrated.