To investigate the vasoconstriction induced by a polymerised bovine haemoglobin solution, Hb-200, in isolated canine arteries.
To investigate the vasoconstriction induced by a polymerised bovine haemoglobin solution, Hb-200, in isolated canine arteries.
Rings of canine saphenous artery, from euthanatized dogs, were mounted between stainless steel wires in Krebs’ solution (95% O2, 5% CO2, 37°C) for isometric tension recording. Following incubation with Hb-200, cumulative concentration response curves to phenylephrine (vasoconstrictor) and acetylcholine (vasodilator) were investigated. Responses to acute addition of Hb-200 were also examined in pre-constricted or pre-dilated arteries. Responses were further studied in the presence or absence of the endothelium, inhibitors of endothelium-dependent vasodilation (L-NAME, charybdotoxin and apamin), an endothelin antagonist (BQ-788) and the antioxidant superoxide dismutase.
Incubation with Hb-200 (0·2 or 2 g/L) significantly enhanced phenylephrine-induced contraction (decreasing half maximal effective concentration, EC50, P=0·0035) and inhibited acetylcholine-induced relaxation (increasing EC50, P<0·0001). Acute addition of Hb-200 (0·2 or 2 g/L) significantly increased tension in pre-constricted arteries (P=0·0059) and reversed relaxation in pre-dilated arteries (P=0·0005). These acute responses were abolished in endothelium-denuded arteries and arteries incubated with L-NAME. Responses to Hb-200 were unaffected by incubation with charybdotoxin and apamin, BQ-788, or superoxide dismutase.
Low concentrations of Hb-200 enhance vasoconstriction in isolated canine saphenous artery, primarily by antagonism of nitric oxide. This effect may be detrimental in some dogs (e.g. those at risk of volume overload) but beneficial in others (e.g. those in septic shock).
There has been considerable interest in the development of haemoglobin-based oxygen-carriers (HBOCs) as an alternative to the transfusion of whole blood and blood products. The only HBOC to gain approval for clinical use in Europe is a veterinary product, haemoglobin glutamer-200 (Hb-200, Oxyglobin; Dechra Veterinary Products Limited). This is an ultrapurified solution of glutaraldehyde polymerised bovine haemoglobin and it is approved for the treatment of anaemia in dogs (European Medicines Agency, 2012). HBOCs have undergone clinical trials in humans and a related product, haemoglobin glutamer-250 (Hemopure; OPK Biotech) has been approved for use in human patients in South Africa. However, safety concerns have been raised following publication of a meta-analysis that showed an increased risk of myocardial infarction and death in humans treated with HBOCs (Natanson et al. 2008).
HBOCs are known to produce vasoconstriction although they vary in their tendency to do so. Early diaspirin cross-linked haemoglobin solutions produced marked vasoconstriction in isolated vessels (Freas et al. 1995, Katušic et al. 1996, Hart et al. 1997) and in vivo (Hess et al. 1993). Polymerised haemoglobins, such as Hb-200, are said to produce less vasoconstriction (Winslow 2000), although increases in vascular tone have been reported (Walton 1996, Driessen et al. 2001, Pawson et al. 2007). Three classes of endothelium-derived vasodilators are involved in the regulation of vascular tone (Fig 1): nitric oxide, prostanoid mediators and endothelium-derived hyperpolarising factors (EDHFs) (Vanhoutte & Mombouli 1996). The vasoconstrictive action of Hb-200 could result from interference with one or more of these vasodilator systems. Antagonism of nitric oxide would appear to be the most important mechanism (Katušic et al. 1996, Hart et al. 1997), although not all studies support this view (Fitzpatrick et al. 2004). Other possible mechanisms include increased release of the vasoconstrictor endothelin (Schultz et al. 1993) and an autoregulatory response to increased oxygen delivery (Rohlfs et al. 1998). Free radicals generated by the auto-oxidation of haemoglobin may contribute to the process by inactivating nitric oxide (Angulo et al. 1996) and/or enhancing endothelin release (Smani et al. 2007).
There are currently no published studies investigating the effects of polymerised bovine haemoglobin, Hb-200, in isolated canine arteries. Therefore the aims of our study were to quantify the vasoconstriction induced by Hb-200 in rings of canine saphenous artery and to investigate, by the use of various inhibitors, the mechanism by which vasoconstriction occurs. Preliminary studies demonstrated a role for nitric oxide and EDHFs, but not prostanoid mediators, in the modulation of tone in the canine saphenous artery (Pawson et al. 2009). Consequently emphasis was placed on inhibitors of nitric oxide and EDHF when investigating the mechanism of vasoconstriction.
The study was approved by the ethics committee of the School of Veterinary Medicine, University of Glasgow. Sections of saphenous artery were collected, with owner consent, from dogs euthanatized, without prior sedation, using pentobarbitone. The dogs varied with respect to age, breed and health status. Reasons for euthanasia included behavioural problems and chronic medical conditions. Vessels were not collected from dogs euthanatized during acute illness. The arteries were cleared of adherent connective tissue and stored overnight in refrigerated oxygen-free Krebs’ solution with the following composition (μM): NaCl 118, KCl 4·8, CaCl2 2·5, MgSO4 1·2, NaHCO3 24 and D-glucose 11.
Within 24 hours of euthanasia, 2·5-mm wide transverse rings of saphenous artery were prepared using a razor slicing device. Arterial rings were mounted under 1-g resting tension between stainless steel wires in 10-mL organ baths containing warm (37°C) oxygenated (95% O2 and 5% CO2) Krebs’ solution. Tension was recorded isometrically using BIOPAC TSD125C isometric force transducers (Linton Instruments) and displayed using ACQKNOWLEDGE 3.8.1. Biopac data acquisition system (Linton Instruments).
Arterial rings were equilibrated for a minimum of 30 minutes, during which time resting tension was re-adjusted to 1 g, if required. After equilibration, vessel viability was assessed by adding phenylephrine. Arteries that generated less than 0·5 g tension to 3 μM phenylephrine were not used further. Acetylcholine, 1 μM, was used to assess endothelial function and relaxation of 25% or more was taken to indicate a viable endothelium.
Acetylcholine chloride, apamin, BQ-788, L-NAME, phenylephrine and superoxide dismutase were obtained from Sigma. Charybdotoxin was obtained from Latoxan. Drugs were dissolved in and diluted with 0·9% sodium chloride (except for BQ-788 that was dissolved in dimethyl sulphoxide). Hb-200 (Oxyglobin®; OPK Biotech LLC) was obtained from Dechra Veterinary Products. Hb-200 comprises a 130 mg/mL solution of Hb-200 with an average molecular weight of 200,000. A 20 g/L solution of Hb-200 was prepared by dilution with Krebs’ solution on the day of the experiment. Hb-200 was used within 24 hours of removing the sealed overwrap, as recommended by the manufacturer.
Firstly, the effect of incubating arterial rings with Hb-200 was investigated. Endothelium intact rings of saphenous artery were pre-incubated for 20 minutes, either alone or with Hb-200 (0·02, 0·2 or 2 g/L). The change in tension during this 20-minute incubation period was recorded. Cumulative concentration–response curves to the vasoconstrictor phenylephrine (1 nM to 100 μM) were then generated. Alternatively following pre-incubation, vessels were contracted with a submaximal dose of phenylephrine, to achieve 50 to 70% of maximum tone, before recording cumulative concentration–response curves to acetylcholine (1nM to 10 μM).
Experiments were also conducted to investigate the more immediate effect of adding Hb-200 to previously constricted or dilated arterial rings. Arterial rings were first contracted using phenylephrine to achieve 50 to 70% of maximum tone (Fig 2a). Once tone stabilised, Hb-200 (0·02, 0·2 or 2 g/L) was added to the bath. After 20 minutes, the change in tension was recorded and compared to a time-matched control (no Hb-200 added). To study the effect on acetylcholine-induced vasodilation, vessels were first contracted with phenylephrine, then relaxed using 1 μM acetylcholine (Fig 2b). As previously the change in tension 20 minutes after addition of Hb-200 (0·02, 0·2 or 2 g/L) was recorded and compared to a control.
To investigate the mechanism of action of Hb-200, the influence of the endothelium was first examined. The endothelium was removed by mounting the arterial rings on a stainless steel wire and then gently rolling the vessel around the wire for 20 revolutions in each direction. Viability was determined as detailed above and lack of relaxation to 1 μM acetylcholine was taken as confirmation that the endothelium had been successfully removed. The response to acute addition of 0·2 g/L Hb-200 was then determined in pre-contracted arterial rings with and without an endothelium.
In this section of the study, the effects of L-NAME, an inhibitor of nitric oxide synthase, and charybdotoxin and apamin, inhibitors of EDHFs, were evaluated. Endothelium intact arterial rings were incubated for 20 minutes, either alone, with L-NAME (100 μM), with charybdotoxin (0·1 μM) and apamin (0·1 μM) or with all three inhibitors (L-NAME 100 μM, charybdotoxin 0·1 μM and apamin 01 μM). Subsequently, the response to acute addition of 0·2 g/L Hb-200 was examined in rings pre-contracted with a submaximal dose of phenylephrine and vessels pre-contracted and then relaxed with 1 μM acetylcholine.
Endothelium intact arterial rings were incubated for 20 minutes, either alone, with the free radical scavenger superoxide dismutase (300 U/mL) or with the endothelin antagonist BQ-788 (1 μM). The response to acute addition of 0·2 g/L Hb-200 was then measured in vessels pre-contracted with phenylephrine or pre-contracted and then relaxed with 1 μM acetylcholine.
Results are expressed throughout as the mean ±standard error of the mean (sem) of the number (n) of separate observations, each from a separate preparation. Contraction in response to phenylephrine is expressed as tension (g). Relaxations are expressed as a percentage of phenylephrine-induced tone. A power calculation demonstrated that to detect a change of at least 10% with a power of 90%, a minimum of six animals would be required in each experiment (http://www.stat.ubc.ca/~rollin/stats/ssize/n2.html). Statistical comparisons were made using Prism (GraphPad). Data were analysed using one-way analysis of variance with Bonferroni's post hoc test for multiple or selected comparisons. A probability (P) less than or equal to 0·05 was considered significant.
Incubation with Hb-200 had a minimal effect on the resting baseline tone of rings of canine saphenous artery. The change in tone recorded over a 20-minute period was −0·10 ±0·03 g, −0·04 ±0·03 g, 0·00 ±0·05 g and 0·08 ±0·09 g for control vessels and vessels exposed to Hb-200 (0·02, 0·2 or 2 g/L) (P=0·1791, n=7).
The effects on phenylephrine and acetylcholine cumulative concentration–response curves are illustrated in Figs 3 and 4, respectively. The two highest concentrations of Hb-200, 0·2 and 2 g/L, enhanced phenylephrine-induced contraction producing small but statistically significant reductions in half maximal effective concentration, EC50 (Table 1). Hb-200 also inhibited acetylcholine-induced relaxation, with 0·2 and 2 g/L concentrations producing a significant increase in EC50 (Table 1).
|Control||Hb-200 0·02 g/L||Hb-200 0·2 g/L||Hb-200 2 g/L||P value|
|Log EC50 (M)||−5·83 ± 0·05||−5·90 ± 0·04||−5·99 ± 0·04*||−6·07 ± 0·04*||0·0035|
|Emax (g)||7·65 ± 0·78||7·24 ± 0·79||7·53 ± 0·86||8·00 ± 0·69||0·9194|
|Log EC50 (M)||−8·00 ± 0·13||−7·74 ± 0·11||−7·20 ± 0·15*||−7·00 ± 0·16*†||<0·0001|
|Emax (%)||−107 ± 13·4||−103 ± 9·7||−76·9 ± 14·8||−71·8 ± 9·8||0·1141|
The effect of acute addition of Hb-200 on the tone of arterial rings contracted with a submaximal dose of phenylephrine is illustrated in Fig 5. Over the 20-minute measurement period control arterial rings lost tension (−0·34 ±0·32 g). However, tension increased in rings exposed to Hb-200 and the change was significant for 0·2 and 2 g/L Hb-200 (P=0·0059). Figure 6 illustrates the effect of acute addition of Hb-200 to vessels first contracted with phenylephrine and then relaxed with acetylcholine. Over the 20-minute measurement period, acetylcholine-induced vasodilation was well maintained in control arterial rings at −79·4 ±8·6% of phenylephrine-induced tone. However, relaxation was significantly impaired by 0·2 and 2 g/L Hb-200 (P=0·0005).
Incubation with the highest concentration of Hb-200 (2 g/L) caused frothing in the tissue bath and in some cases this led to a gradual decline in the amount of Krebs’. To avoid this complication a concentration of 0·2 g/L Hb-200 was used for subsequent experiments to investigate the mechanism of action.
Figure 7 illustrates the effect of endothelium removal on the acute response to Hb-200. In the absence of a viable endothelium Hb-200 did not produce significant vasoconstriction.
Hb-200 induced a significant increase in tension in pre-contracted control vessels (P=0·0001, Fig 8) and this increase was not affected by incubation with charybdotoxin and apamin. In contrast, incubation with L-NAME virtually abolished the response to Hb-200. In the presence of all three inhibitors, i.e. L-NAME, charybdotoxin and apamin, Hb-200 appears to induce some contraction but this was not statistically significant.
In vessels pre-contracted with phenylephrine and then relaxed with acetylcholine (Fig 9), charybdotoxin and apamin did not affect either the relaxant response to acetylcholine or the Hb-200-induced reversal of this response. However, L-NAME, alone or in combination with charybdotoxin and apamin, inhibited acetylcholine-induced relaxation, and on addition of Hb-200 no further change in tone occurred.
The increase in phenylephrine-induced tension in response to Hb-200 was not significantly altered by incubation with superoxide dismutase or the endothelin antagonist BQ-788 (Fig 10). In a second experiment, neither superoxide dismutase nor BQ-788 impaired the response to Hb-200 in vessels previously relaxed with acetylcholine (Fig 11). If anything, the response to Hb-200 was enhanced when compared to control vessels although the response in control vessels was less than previously observed (compared to Fig 9).
Hb-200 enhanced phenylephrine-induced contraction and impaired acetylcholine-induced vasodilation in rings of canine saphenous artery. These responses are qualitatively similar to those reported previously in isolated vessels, using earlier HBOCs (Katušic et al. 1996, Muldoon et al. 1996, Hart et al. 1997) and using Hb-200 in isolated rat aorta (Pawson et al. 2007).
Differences in the magnitude of response to HBOCs are apparent on reviewing the literature and may reflect differences between species, the size and location of the vessel studied and the vasoactive properties of individual HBOCs. Hart et al. (1997) examined the effects of a diaspirin cross-linked haemoglobin in a range of arteries from the dog and rat. Where equivalent arteries were compared in both species the response tended to be greater in rat arteries. A greater magnitude of response was also observed in large conducting arteries, such as the rat aorta and canine pulmonary artery. When different HBOCs are compared molecular size is a key determinant of vasoactivity (Sakai et al. 2000). Large molecules are less able to diffuse across the endothelial barrier, thereby limiting the extent of nitric oxide scavenging. On this basis, polymerised haemoglobins such as Hb-200 have been assumed to be less vasoactive than cross-linked haemoglobins (Winslow 2000). However, Hb-200 is very heterogeneous with respect to molecular weight, containing a relatively high proportion of smaller haemoglobin aggregates (Soma et al. 2005), and this may enhance its vasoactivity.
Dose-dependency in the response to Hb-200 is evident in the results of this study and has been observed previously with other types of HBOC (Hart et al. 1997). The greatest magnitude of effect was observed with the highest concentration of Hb-200 used, i.e. 2 g/L. However, this concentration produced frothing in the tissue bath and so a concentration of 0·2 g/L Hb-200 was used in subsequent experiments. The concentrations studied are low when compared to anticipated clinical concentrations. According to the manufacturer's information (National Office of Animal Health 2013), infusion of 15 mL/kg Hb-200 will result in an immediate post-infusion plasma concentration of 20 to 25 g/L. This is 10 times greater than the highest concentration evaluated in this study.
Removal of the endothelium from the arterial rings abolished the response to Hb-200 confirming that it acts via an endothelium-dependent mechanism. L-NAME, an inhibitor of nitric oxide synthase, significantly impaired the response to Hb-200 in both constricted and relaxed arterial rings. This finding confirms that interference with nitric oxide is integral to the vasoconstrictive action of Hb-200 in isolated canine vessels, as it is for diaspirin cross-linked haemoglobin solutions (Katušic et al. 1996, Muldoon et al. 1996). This may explain the greater response to HBOCs observed in large conducting arteries, since nitric oxide production is greatest in such vessels (Levenson et al. 1985). Alternative vasodilators, such as EDHFs, are thought to be more important modulators of tone in smaller vessels. Charybdotoxin and apamin inhibit calcium-activated potassium channels and thereby prevent the action of EDHFs (Bryan et al. 2005). Incubation with these inhibitors did not significantly alter Hb-200-induced increases in vascular tone, leading to the conclusion that EDHF is not involved in the response to Hb-200 in these arteries.
Several groups have demonstrated a potential role for the vasoconstrictor endothelin in the response to HBOCs (Schultz et al. 1993, Ledvina et al. 1999, Smani et al. 2007). Endothelin is a potent endogenous vasoconstrictor produced by the vascular endothelium. Two endothelin receptor types are recognised: type A receptors (ETR-A) are confined to smooth muscle cells, whereas type B receptors (ETR-B) are also located on endothelial cells. There is evidence that the ETR-B receptor is most important in the response to HBOCs (Heller et al. 1998), consequently an ETR-B antagonist, BQ-788, was used in this study.
Endogenous vasodilator and vasoconstrictor systems interact in complex ways and inhibition of nitric oxide has been shown to increase endothelin release and activity (Boulanger & Lüscher 1990, Lerman et al. 1992). It has been postulated that by inactivating nitric oxide, HBOCs promote endothelin synthesis and that this in turn contributes to vasoconstriction (Lin et al. 2001). If this is correct, incubation with BQ-788 would be expected to attenuate the response to HBOCs. This has been demonstrated by some investigators (Smani et al. 2007) but in this study, BQ-788 had no significant effect on the response to Hb-200. Despite this finding, a role for endothelin should not be discounted without also examining the effects of an ETR-A receptor antagonist.
HBOCs can alter redox status (Motterlini et al. 1995). The Fe2+ in haem undergoes auto-oxidation to Fe3+, a process that generates harmful free radicals, notably superoxide anion (O2•−) which is known to directly inactivate nitric oxide and thus cause vasoconstriction. Smani et al. (2007) demonstrated an increase in production of O2•− during exchange transfusion of a conjugated HBOC (dextran-benzene-tetracarboxylate-haemoglobin) to guinea pigs and proposed that free radicals are important mediators of HBOC-induced vasoconstriction. If this is the case, superoxide dismutase, an antioxidant enzyme that catalyses the dismutation of O2•− to oxygen and hydrogen peroxide, might be expected to diminish HBOC-induced vasoconstriction. However, superoxide dismutase did not attenuate the response to Hb-200 in this study, providing no evidence that free radicals are responsible for Hb-200-induced vasoconstriction.
There are limitations to this study. The number of animals in each group is small, ranging from 6 to 9. A power calculation demonstrated that a minimum group size of 6 would be required but given the diverse population of dogs from which vessels were collected, i.e. different with respect to bodyweight, breed, gender, age and state of health, a larger group size would have provided more statistical power. However, collecting sufficient tissue proved to be challenging. Administration of sedatives before euthanasia, specifically acepromazine, affected the contractile function of vessels collected and so only tissue from unsedated dogs could be used. In addition, overnight storage in cold Kreb's solution may have diminished contractile and relaxant responses (Stanke-Labesque et al. 1999).
Charybdotoxin, has been shown to induce spontaneous contractile activity in some arterial preparations (Gokina et al. 1996). Unstable vascular tone and spontaneous contractions were occasionally observed in this study. This activity may account for the apparent response to Hb-200 recorded in arteries incubated with L-NAME, charybdotoxin and apamin (Fig 8), a response that is otherwise difficult to explain.
The findings of this and previous studies highlight the nitric oxide scavenging activity of Hb-200. This activity should be considered when Hb-200 is used clinically. Vasoconstriction is the most obvious consequence and may contribute to the development of volume overload reported in some patients, especially cats (Gibson et al. 2002); although the fluid's high colloid osmotic pressure is also undoubtedly important. In other patients, such as those in septic shock, this activity may be beneficial as nitric oxide excess is known to contribute to the inappropriate vasodilation and refractory hypotension observed (Kirkebøen & Strand 1999). However, nitric oxide is much more than a vasodilator (Jin & Loscalzo 2010), modifying the function of leucocytes and platelets and notably inhibiting platelet aggregation. Thus a nitric oxide scavenger might also be expected to promote platelet activity, an effect that could be detrimental in some individuals. Clearly, further studies are required to investigate the wider implications of nitric oxide scavenging.
In conclusion, this study demonstrates that low concentrations of Hb-200 significantly enhance vasoconstriction in isolated rings of canine saphenous artery and that this effect is mediated primarily by antagonism of nitric oxide. Neither endothelin, acting via ETR-B receptors, nor free radical generation contributed significantly to the activity of Hb-200 under the conditions of this study.
None of the authors of this article has a financial or personal relationship with other people or organisations that could inappropriately influence or bias the content of the paper.
The authors gratefully acknowledge BSAVA Petsavers for funding this project and would like to thank Mr Ian Gibson for his technical support.