ACD Blood Research Articles

A High Parameter Flow Cytometric Assay to Identify Human Myeloid Derived Suppressive Cells

Abstract The concept of myeloid derived suppressive cells (MDSCs) has emerged in recent years as a group of myeloid cells that pocess potent immunosuppressive activity. These cells expand under pathologic conditions and suppress the immune functions of T cells, dendritic cells, macrophages, and natural killer cells. They have been shown to be involved in the pathological states of cancer, autoimmune diseases, and infections, such as COVID-19. MDSCs are subset into PMN (polymorphonuclear)- and M (monocytic)-MDSCs. Phenotypic markers of MDSCs have been elucidated in recent years. Here Cytek® developed a 14-color, 16-marker, single-tube flow cytometric assay to analyzed MDSCs. The panel includes markers to exclude lymphoid lineage cells, and markers to identify PMN-MDSC and M-MDSC populations, such as CD15, CD14, CD16, CD66b, CD11b and HLA-DR. In addition, PMN- and M-MDSC specific expression markers such as Lox-1, CXCR-1, DR5 are also included. Fresh whole blood samples were stained with erythrocyte lyse wash method and the samples were analyzed on a Cytek® Aurora or 3-laser Northern Light TMinstrument. The panel performance was evaluated on blood samples collected in EDTA, Heparin, ACD, and Cytochex blood tubes as well as peripheral blood mononuclear cell (PBMC) samples. Assay precision was assessed by testing on 12 donor samples in triplicates on three consecutive days. Targeted cell populations are well identified in peripheral blood and PBMC samples with good resolution and assay precision with %CV less than 25% for all cell populations of interests. This 14-color 16-marker preoptimized flow cytometric assay can be used in clinical research and drug discovery in cancer, autoimmune diseases, infections and other inflammatory diseases.

Accurate enumeration of leukocyte subsets in peripheral blood with a 15-color immunoprofiling panel on a Cytek ®spectral flow cytometer

Abstract A white blood cell (WBC) differential is routinely performed on automated hematology analyzers as a general health check to evaluate immune status of patients. The recent advances in high parameter flow cytometry, especially full spectral flow cytometry, enable the analysis of more comprehensive leukocyte subsets to meet the increased demands in clinical research. We designed a 15-color, single-tube panel to identify and enumerate all major subsets of leukocytes on a Cytek full spectral flow cytometer. Human peripheral blood collected in K2EDTA, Heparin, ACD and Cyto-chex blood tubes were stained with 15 cFluor ®conjugated mAbs with a Lyse No Wash method and analyzed on a Cytek spectral flow cytometer. The performance of this assay was evaluated side by side with BD Multitest™ 6-color TBNK assay on BD FACSCanto II system on 13 healthy donor samples. Assay precision was examined on 12 donor samples in triplicates on three consecutive days. This 15-color panel identifies more than 20 leukocyte subpopulations: including neutrophils, eosinophils, basophils, lymphocytes, hematopoietic stem cells, monocyte subsets, T cell subsets, B cells, NK cell subsets. The results are comparable to that from BD 6-Color TBNK assay with R 2=0.961, 0.986, 0.940, 0.954, and 0.930 for CD3+, CD4+, CD8+ T, B and NK cells respectively. The correlations for all subsets are statistically significant with P < 0.001. The coefficient of variations (CV) of intra assay precision for all cell populations of interest are less than 25%. Cytek’s single-tube 15-color pan leukocyte panel demonstrated an effective, high sensitivity flow cytometry approach that can be used for monitoring immune cell subsets in peripheral blood in discovery and translational research.

Cation Flux and Electrolyte Composition of Frozen-Deglycerolized Blood

Cation flux and the extracellular and erythrocyte concentrations of sodium, potassium, calcium, magnesium, chloride, and acid soluble phosphorus were measured in fresh blood collected in various anticoagulant solutions, ACD blood stored at 4 C, and in frozen-deglycerolized blood stored at -80 to -120 C. Considerable variations in concentrations of extracellular and erythrocyte electrolytes were encountered due to certain characteristics of the anticoagulants employed and to the effects of storage, temperature, pH, and extracellular electrolyte concentration on erythrocyte electrolyte concentration. In contrast to ACD blood stored at 4 C, frozen-deglycerolized blood 1 day to 32 months old had slightly elevated erythrocyte concentrations of sodium and chloride and normal erythrocyte concentrations of potassium and phosphorus. Both ACD and frozen-deglycerolized blood had normal sodium and potassium fluxes. These data show that prolonged freezing and glycerolization of erythrocytes do not significantly impair the active transport processes responsible for maintenance of cellular electrolyte concentration.

P105 An accelerated method for reverse sequence-specific oligonucleotide probe-based HLA typing

Turnaround time (TAT) for HLA typing is critical for transplantation. We sought to establish a faster protocol for low-resolution HLA typing using the rSSOP methodology. The LABType™ rSSOP method (One Lambda) was altered to reduce hybridization time. Modifications included centrifugation and incubation duration reductions, and reagent volume adjustments. Single parameters were first changed and tested. Successful modifications were used in the next round of tests. All modifications were tested with a set of PCR amplicons prepared from three samples, which contained previously HLA-typed high quality DNA. Following a series of rSSOP typings for these samples at HLA-A, B, C, DRB1, and DQB1/DQA1 loci, an accelerated protocol was established. Discovery phase concluded with protocol validation at all loci including HLA-DRB3/4/5 and DPB1/DPA1. Full validation confirmed intra-operator, inter- and intra-assay, as well as multiple DNA source reproducibility. DNA was extracted from different samples including ACD blood (n = 4), EDTA blood (n = 4), cord segment (n = 5), buccal swab (n = 5) and mouth-wash samples (n = 4). The original centrifugation of 3000 rpm for 2 min (total 2 min 22 s) was changed to a quick spin (up to 3000 rpm; 33 s), reducing each centrifugation step by 1 min 50 s. Additionally, one centrifugation and one transfer of plate step were eliminated. The duration of incubation time was changed for denaturation (from 10 min to 2.5 min), hybridization (from 15 min to 5 min), and streptavidin-PE (SAPE, from 5 min to 2.5 min). The reagent volume modifications includes the reduction of washing buffer (from 100 μl to 50 μl per well) and increase in SAPE (from 25 μl to 50 μl per well). Compared to the original protocol, the accelerated protocol reduced the hybridization time approximately 30 min. This method allowed us to obtain consistent and accurate typing results for all tested HLA loci among different operators, intra- and inter-assays, and a variety of DNA samples. Fulfilling requests for faster TAT may be achieved by adaptation of current methodologies, which may help improving patient management. The accelerated rSSOP protocol is an effective method for reducing testing time, while still giving robust results for HLA class I and class II typing. (RLU and JJX contributed equally).

Functional stability of ACD anticoagulated blood samples stored for 72 hours at room temperature.

Abstract Characterization of the immune system in remote locations is challenging, requiring technical expertise and specific analytical equipment. Immune functional assessments require live cells, and the window of viability following blood donation is limited. To test the ability of ACD blood collection tubes to expand this window of viability, cellular function was analyzed over 4 consecutive days, beginning on the date of blood collection. T-cell activation, characterized by CD69 and CD25 expression, was measured following a 24-hour stimulation with soluble anti-CD3 and anti-CD28 antibodies (CD3/CD28) or with staphylococcus enterotoxins A and B (SEA+SEB). Mitogen stimulated cytokine profiles were determined following a 48-hour stimulation with CD3/CD28, phytohemagglutinin and ionomycin (PMA+I), or lipopolysaccharide (LPS). T-cell activation following stimulation with SEA+SEB was not altered by the processing delay; however, following stimulation with CD3/CD28, T-cell activation increased over the 72 hour period. The processing delay did not affect cytokine production following CD3/CD28 stimulation. Following stimulation with LPS and PMA+I, results were varied. While some cytokine responses were unaltered following the processing delay, others were impaired. These findings confirm the general viability of whole blood stored at room temperature (68 to 72°F) for 48 to 72 hours in ACD tubes and suggest many phenotypic and functional assays are feasible on blood samples aged as described. Investigators using this protocol should confirm their specific culture technique and measured output. This protocol may assist studies of subjects in remote locations with no immediate access to laboratory instrumentation.

P117 Swapping the swab for bmt confirmatory typing

Aim Our lab switched from ACD blood source to buccal swab source for confirmatory typing of HSCT (Hematopoietic Stem Cell Transplantation). The buccal specimen source met our goals for a second sample source and determining true homozygosity versus allele drop out. We will discuss implementing the change and a case of HLA-DQB1 drop out. Methods Primary typing of HLA low resolution (Sequence Specific Oligonucleotide, One Lambda) and high resolution typing (SeCore, One Lambda) are completed on ACD blood. Implementation of the buccal swab in our lab included validation of a new extraction kit, standard operating procedural (SOP) changes, test code/resulting changes, and training. Results Validation and integration of the EZ1 DNA Tissue kit (Qiagen) were simple to complete. SOP changes and resulting added the process to include referencing the white blood count (WBC) within the week of the blood collection for the primary typing. Primary typing results homozygous at any loci accompanied with a WBC great than 10 × 10(9)/L, triggers a closer look with the buccal sample. An obstacle we encountered was the collection of the buccal leading to low amounts of DNA for testing. We provided feedback to our collection groups and have observed fewer re-collections the past year. A difference observed between blood and buccal swab was for a patient’s DQB1 locus. SSO/SeCore from blood sample demonstrated homozygous DQB1 ∗ 05. DRB1 resulted ∗ 01, ∗ 04; therefore, we more commonly expected a second DQB1 antigen other than ∗ 05. The WBC result from the blood collection date was 57.7 × 10(9)/L. We obtained DNA from buccal sample and re-tested. Assays now cleanly demonstrated heterozygous DQB1 ∗ 05, ∗ 03. Conclusion Switching sample sources created process changes for the laboratory and the collection group. Training and process flow in the lab was successful. Feedback to the collection group was needed to ensure swabbing was adequate for a good DNA yield for testing. We feel the testing algorithm provides assurance of accurate test results for HSCT patients.

1016-LBP: RosetteSep HLA – An exceptional method for isolation of lymphocytes from peripheral blood

Aim The isolation of a sufficient quantity of lymphocytes from deceased donor peripheral blood can be problematic. As an OPO HLA lab that serves six different transplant centers, we routinely perform both complement dependent cytotoxicity (CDC) and flow cytometric crossmatching (FCXM) for 15 potential renal recipients plus additional non-renal recipients. Historically, we have used T/B Lympho-Kwik to isolate lymphocytes for FCXM, but this reagent is no longer available. As a replacement method, we adopted a protocol using CD14 and CD15 magnetic beads to negatively select non-lymphocyte cells. However, the lymphocyte yield and purity was frequently insufficient, so we sought to validate a method using RosetteSep HLA Total Lymphocyte Enrichment Cocktail. Methods RosetteSep contains bispecific antibodies which crosslink unwanted cells to RBCs, forming rosettes, which are removed by centrifugation. Buffy coats were pulled from ACD blood tubes and lymphocytes were isolated by centrifugation over IsoPrep and negative selection with CD14/CD15 Dynabeads or RosetteSep. The isolated lymphocytes were counted on a hemacytometer and either used directly for FCXM or further separated by positive selection of B cells using Class II Dynabeads for CDC testing. Results The average cell yield for the RosetteSep isolations was nearly double that of the CD14/CD15 isolations, lymphocyte purity was >95%, and there was an increase in the ratio of B cells to T cells. The viability and CDC reaction of the B cells was comparable between the two methods. The results of the FCXM were also comparable between the two methods, both in terms of T and B cell MFI with negative serum (NS) and T and B cell MCS with patient sera. The increased background on B cells with NS historically seen when using Lympho–Kwik was not observed with RosetteSep. Conclusions Our method using RosetteSep is cost efficient, fast, easy to perform, and significantly increases the yield of lymphocytes compared to CD14/CD15 magnetic bead negative selection.

Blood group phenotype frequencies in blood donors from a tertiary care hospital in north India

BackgroundKnowledge about the frequency of red blood cell-antigen phenotypes in a population can be helpful in the creation of a donor data bank for the preparation of indigenous cell panels and for providing antigen-negative compatible blood to patients with multiple alloantibodies.MethodsABO and RhD blood grouping was performed on 9,280 continuous voluntary and replacement donors. For other rare blood groups, 508 ACD blood samples were obtained from the donors at the Blood Bank of the Department of Transfusion Medicine, All India Institute of Medical Sciences (AIIMS), New Delhi, India. Blood group antigens were determined by tube method using anti-sera (Bio-Rad, USA), and the phenotype frequencies were expressed as percentages.ResultsGroup B (37.39%) was the most common, followed by group O (31.85%). R1R1 and rr were the most common phenotypes amongst Rh positive and Rh negative groups, respectively. A rare phenotype R2Rz was found in one donor. For Kidd and Duffy blood group systems, Jk (a+b+) and Fy (a+b+) were the most common phenotypes (46.06% and 48.03%, respectively). The most common phenotypes for MNSs, Lu, and Kell blood groups were M+N+, S-s+, Lu (a-b+), and K-k+, respectively. A very rare case of Fy (a-b-) and Jk (a-b-) was found in a single donor.ConclusionThis study is the first small step to create a rare donor data bank and to prepare indigenous cell panels to provide compatible blood to all multi-transfused alloimmunized patients.

Volume of RBCs, 24- and 48-hour posttransfusion survivals, and the lifespan of (51)Cr and biotin-X-N-hydroxysuccinimide (NHS)-labeled autologous baboon RBCs: effect of the anticoagulant and blood pH on (51)Cr and biotin-X-NHS elution in vivo.

The RBC injury that occurs during collection of the first few milliliters of blood into the pH 5.0 ACD (NIH, Formula A) is referred to as the lesion of collection. The RBC injury was evaluated by labeling the ACD RBCs with (51)Cr and measuring the 24-hour posttransfusion survival. The effect of the acidification of ACD blood on the in vivo elution of (51)Cr from the RBC has not been reported. Baboon blood was collected in heparin and in ACD and CP2D at different ratios of blood to anticoagulant. ACD blood with a pH of 5.7 to 6.9 was labeled with (51)Cr. Heparinized blood with a pH of 7.4 was labeled with biotin-X-N-hydroxysuccinimide (NHS). ACD blood with a pH of 5.9 was labeled with both (51)Cr and biotin-X-NHS. The RBC volumes, 24- and 48-hour posttransfusion survivals, and lifespans (T50) were measured. The RBC volume of ACD blood with a pH ranging from 5.7 to 6.9 was not affected by (51)Cr labeling. (51)Cr-labeled ACD blood with a pH of 6.9 had an RBC volume that was significantly greater than that seen in heparinized blood with a pH of 7.4 labeled with biotin-X-NHS. In vivo elution of (51)Cr from the RBCs prepared from the ACD blood with a pH of 5.7 to 6.9 and in vivo elution of biotin-X-NHS from RBCs prepared from ACD blood with a pH of 5.9 were associated with reductions in 24- and 48-hour posttransfusion survivals and T50. The anticoagulant and the pH of the medium in which the RBCs were labeled with (51)Cr or biotin-X-NHS affected in vivo elution of the label from the RBCs and may reduce posttransfusion survival values.

Heparin-coated oxygenators significantly reduce contact system activation in an in vitro cardiopulmonary bypass model.

Over the past decade our group has shown that the contact system of blood is activated in cardiopulmonary bypass (CPB), that heparins enhance this activation and that aprotinin reduces both this activation and blood loss in CPB. We have developed an in vitro CPB model to assess the effects of added components to blood and new components in the artificial devices of CPB. In the present study we have compared membrane oxygenators with or without heparin-coated surfaces under identical conditions in the CPB model. Recalcified ACD blood was circulated in a closed system for 90 min at 28 degrees C. Blood samples were taken at various times during circulation. 4 IU/ml heparin was used with the non-coated oxygenators; no heparin was used in the coated system. Heparin levels were measured in the plasma together with various contact system components. Haemolysis, platelet count, platelet factor 4 and alpha 1-proteinase inhibitor-PMN elastase complexes, were also determined. No heparin was detected during the period of recirculation in samples from the coated oxygenators, showing the excellent adhesive quality of the heparin coating. In keeping with a significant greater fall in the platelet count in non-coated vs coated oxygenators (mean [+/- SD] final counts of 170 +/- 50 x 10(9)/l and 97 +/- 34.2 x 10(9)/l respectively after 90 min circulation), platelet factor 4 levels were significantly higher (682.9 +/- 187.3% and 95.8 +/- 46.5% of the initial value respectively).(ABSTRACT TRUNCATED AT 250 WORDS)

In‐Vitro Assessment of Centrifugal Pumps for Ventricular Assist

Currently two major types of centrifugal pumps are commercially available for ventricular assist: the Biomedicus-cone (Group I) and the Centrimed-impeller pump (now Sarns 3M) (Group II). To compare them for blood trauma and hemolysis, an in-vitro experiment was designed with a Stöckert roller pump as a standard control (Group III). The in-vitro circuit was constructed consisting of a pump head, electromagnetic flow probe, polyvinyl chloride tubing and a reservoir, identical for all groups. Human ACD blood was used for priming and was circulated with a flow rate of 2 L/min for 24 h. Blood samples were taken at 0, 1, 3, 6, 12, and 24 h and zero control values were subtracted from the resulting data per time interval. Among the 16 parameters studied, a highly significant difference in favor of Group I was found for glutamate oxalacetate transaminase (GOT) and lactate dehydrogenase (p less than 0.0001) and for the free plasma hemoglobin (p less than 0.0001) after 12 and 24 h, respectively. The hemolytic index (Allen) again was lowest for group I in contrast to Groups II and III (0.012 versus 0.060 and 1.70) after 24 h. All other parameters studied did not render significant differences between the systems tested. The authors conclude that the Biomedicus pump currently is the least traumatic centrifugal pump for ventricular assist.

Ca 2+- and Mg-ATP-dependent shape change of human erythrocyte ghosts and Triton shells

Human erythrocyte membranes (ghosts) prepared from fresh blood changed in shape from spherical to crenated, when suspended in 10 −7–10 −6 M Ca 2+-EGTA buffers. Although the ghosts from long-stored ACD blood (10 weeks) were less sensitive to 10 −7–10 −6 M Ca 2+, the ghosts obtained from this blood after it had been preincubated with adenine and inosine for 3 h at 37 °C were highly sensitive to Ca 2+. When these highly sensitive ghosts were incubated in 10 mM Tris-Cl buffer (pH 7.4) or 1 mM MgCl 2 (pH 7.4) at 0 °C, they gradually lost Ca 2+ sensitivity within 60 min, but they recovered Ca 2+ sensitivity again after re-incubation with 2 mM Mg-ATP for 20 min at 37 °C followed by washing with 1 mM MgCl 2 (pH 7.4). The shape of these highly Ca 2+-sensitive ghosts immediately changed from crenate to disc on addition of 1 mM Mg-ATP even at 6 °C in the presence of 10 −7–10 −6 M Ca 2+. A similar shape change was also observed when ghosts treated with 0.5% Triton X-100 (Triton shells) were used. Triton shells from fresh blood ghosts or from long-stored blood ghosts which had been preincubated with 2 mM Mg-ATP for 20 min at 37 °C shrank immediately in the presence of 10 −6 M Ca 2+ and then swelled on addition of 1 mM Mg-ATP. The specificity to ATP and the dependency on ATP concentration are in agreement with those of the ghost shape change at step 2 (Jinbu, Y. et al., Biochem biophys res commun 112 (1983) 384–390) [18]. These results suggest that cytoskeletal protein phosphorylation enhances sensitivity to Ca 2+ and induces erythrocyte shape change in the presence of physiological concentrations of ATP and Ca 2+.

Changen in serum ionized calcium following blood transfusion and calcium gluconate administration

We measured the serum ionized calcium concentration (Ca++) in 27 patients who received ACD blood transfusion ranged from 400ml to 4200ml during general anesthesia. Ca++ was measured prior to transfusion, at various intervals and at the end of transfusion.Ca++ decreased immediately after the transfusion, from 2.13±0.14mEq/l at the baseline to 1.92±0.20 mEq/l (p<0.05) following the 200ml of transfusion and tended to decrease with additional transfusion up to 600ml of blood. The lowest value of Ca++ was 1.78±0.28mEq/l at the 2400-4200ml of transfusion. All values were significant (p<0.05) to baseline, but not significant between each other. We observed no correlation between the rate of transfusion and the decline of Ca++.8 patients who received 1350ml of transfusion on the average, were administered 5ml of 8.5% calcium gluconate after the termination of trantfusion. Ca++ increased from 1.72±0.16mEq/l to 2.06±0.21mEq/l (p<0.05) 5min after the administration, and returned to the normal value.

保存ヒト赤血球の加齢による酵素活性変化

Human red cells were obtained from ACD blood, and named the upper one-thirds as a top fraction and lower one-thirds as a bottom fraction after low speed centrifugation. Each sample was stored in a refrigerator (4°C) and in a freezer (-30°C) during 4 weeks, respectively.The activities of xylulokinase, hexokinase, transketolase, glucose-6-phosphate dehydrogenase and lactate dehydrogenase were measured by enzymatic methods on each sample.All activities of enzymes decreased more in the bottom fraction than in the top fraction.It was suggested that the decrease of activities in the fraction which finally showed some difference with cell age in vivo were in parallel with those of in vitro.

Differential coating of human red blood cells with C4 or C3 in a low ionic strength medium.

An EDTA-containing, low ionic strength medium, pH 5.1, can be used for incubation of blood to obtain coating of red blood cells with C4 or C3. The coating achieved depends on incubation temperature (4 or 37 degrees C) and on whether or not CaCl2 or MgCl2 is added to the medium. The optimal concentration of EDTA appears to be 2 mM under the present conditions. This is equimolar to the final concentration of the added CaCl/MgCl2. The choice between CaCl and MgCl2 depends on whether blood without anticoagulant or ACD blood is employed. Coated cells keep satisfactorily for 5 weeks at 4 degrees C in Alsever's solution as well as after freezing and thawing. C3-coated cells can be converted to C3d-coated ones by incubation in normal, compatible, EDTA-containing serum at 37 degrees C.

Alterations of saponin hemolysis during storage of ACD blood.

Abstract. A dynamic hemolysis curve was examined on stored ACD blood by a new instrument (mean cell volume or MCV continuous analyzer) during treatment with either saponin or with hypotonic medium. Several changes of the dynamic hemolysis curves were recorded during storage. With saponin hemolysis, these changes appeared earlier and were more marked than those of hypotonic hemolysis.

Differential Coating of Human Red Blood Cells with C4 or C3 in a Low Ionic Strength Medium

An EDTA-containing, low ionic strength medium, pH 5.1, can be used for incubation of blood to obtain coating of red blood cells with C4 or C3. The coating achieved depends on incubation temperature (4 or 37°C) and on whether or not CaCl(2) or MgCl(2) is added to the medium. The optimal concentration of EDTA appears to be 2 mM under the present conditions. This is equimolar to the final concentration of the added CaCl(2)/MgCl(2). The choice between CaCl(2) and MgCl(2) depends on whether blood without anticoagulant or ACD blood is employed. Coated cells keep satisfactorily for 5 weeks at 4°C in Alsever’s solution as well as after freezing and thawing. C3-coated cells can be converted to C3d-coated ones by incubation in normal, compatible, EDTA-containing serum at 37°C.

Alterations of Saponin Hemolysis during Storage of ACD Blood

A dynamic hemolysis curve was examined on stored ACD blood by a new instrument (mean cell volume or MCV continuous analyzer) during treatment with either saponin or with hypotonic medium. Several changes of the dynamic hemolysis curves were recorded during storage. With saponin hemolysis, these changes appeared earlier and were more marked than those of hypotonic hemolysis.

Met-form hemoglobins in long-term stored ACD blood.

On passing ACD blood stored for a year at 4 degrees C through a column of Sephadex G-75, the surviving erythrocytes were separated from hemolyzed cells and plasma. About 18% of stored erythrocytes survived a year of storage, and were spherocytic. The concentration of total ferric hemoglobins in the surviving erythrocytes was about 75% of that in the supernatant. The met-form hemoglobins in the erythrocytes and hemolytic solutions were further analyzed by isoelectric focusing electrophoresis, and (alpha 2+ beta 3+)2 and (alpha 3+ beta 2+)2 valency hybrids were found in addition to methemoglobin. Since the activity of NADH cytochrome b5 reductase was extremely low in the surviving erkythrocytes, the difference in proportion of ferric hemoglobins (valency hjybrids + methemoglobin) in these two fractions might be due simply to the difference in pH between intracellular and extracellular fluids, which is largely reflected in the oxidation rate of hemoglobin. The course of hemoglobin oxidation seems to be through hemoglobins such as (alpha 2+ beta 3+)2 and (alpha 3+ beta 2+)2 to methemoglobin.

Met-Form Hemoglobins in Long-Term Stored ACD Blood

On passing ACD blood stored for a year at 4°C through a column of Sephadex G-75, the surviving erythrocytes were separated from hemolyzed cells and plasma. About 18% of stored erythrocytes survived a year of storage, and were spherocytic. The concentration of total ferric hemoglobins in the surviving erythrocytes was about 75% of that in the supernatant. The met-form hemoglobins in the erythrocytes and hemolytic solutions were further analyzed by isoelectric focusing electrophoresis, and (α^2+β^3+)(2) and (α^3+ß^2+)(2) valency hybrids were found in addition to methemoglobin. Since the activity of NADH cytochrome b(5) reductase was extremely low in the surviving erythrocytes, the difference in proportion of ferric hemoglobins (valency hybrids + methemoglobin) in these two fractions might be due simply to the dilference in pH between intracellular and extracellular fluids, which is largely reflected in the oxidation rate of hemoglobin. The course of hemoglobin oxidation seems to be through hemoglobins such as (α^2+β^3+)(2) and (α^3+ß^2+)(2) to methemoglobin.