The Long Road to Ithaca: A Physiologist’s Journey

Prologue

It was my great honor to be asked to deliver the Walter B. Cannon Lecture during the 2024 American Physiological Summit meeting. Dr. Cannon, was a giant in our field and a person to emulate. After graduating from Harvard Medical School, he joined the department of Physiology as an instructor. Then, in 1906 he succeeded his mentor, Dr. Bowditch, as the Higginson Professor and Chair of the Department of Physiology at Harvard Medical School, a position he held for the next thirty-six years.

Dr. Cannon played a seminal role in establishing Physiological Reviews, currently the most cited journal in physiology, and served as president of the American Physiological Society from 1914–1916. He coined the term “fight or flight” to describe an animal’s response to threats and the concept of Homeostasis, the “Physiological mechanisms that maintain constant the variables related to the internal milieu” [1]. He published numerous articles and books and had a mountain peak named after him by the United States Geological Survey. He was elected a Member of the American Academy of Arts and Sciences (1906) and the National Academy of Science (1914). He was an outstanding mentor having benefited from the mentoring of Dr. Henry Pickering Bowditch, who suggested to Mr. Cannon, a 2nd year medical student at the time, to investigate the mechanisms of swallowing using x-rays, which resulted in his first publication in the American Journal of Physiology in 1898 [2]. During the first World War I, he joined a Harvard medical unit in Europe to study and combat traumatic shock [3] and helped save many lives of wounded soldiers. He was the consummate physician-scientist, an outstanding mentor and teacher, a prolific writer, and a humanitarian.

The title of my lecture was “The Long Road to Ithaca: A Physiologist’s Journey.” The title is based on a poem entitled “Ithaca,” written by the Greek poet C. P. Cavafy, who recounts the 10 year travels of Ulysses, from Troy to Ithaka, a small island in the Ionian sea. In this poem, he stresses that Odysseus had to overcome many obstacles to survive this long journey; however, he enjoyed the adventures and the knowledge he gained by overcoming the many obstacles which lifted his spirit and transformed him into a wiser person.

I thus used Ithaca as a metaphor to recount my adventures, challenges and joys during my fifty years as a basic and translational scientist, teacher, mentor, and editor. Like Odysseus, I encountered myriad of problems: editors rejected my papers, NIH study sections found my grant applications wanting, senior scientists criticized my findings (“Young man, Startling would be turning in his grave if he listened to your presentations;” APS meeting, 1976). I also survived three bouts of cancer and a catastrophic heart attack that left me without a heart-beat for 15 min. I am still alive thanks to my wife, Dr. Lee Ann Riesenberg, who gave me CPR till the paramedics arrived. But I also have experienced the thrill of contributing to scientific knowledge, the satisfaction of watching my mentees develop into independent scientists, the excitement of teaching respiration physiology to medical and professional students, and the pleasure of being of service to my discipline by serving as Editor of the American Journal of Physiology-Lung Cellular and Molecular Physiology and of Physiological Reviews. I have been a member of the American Physiological Society since the early 70’s; I have benefited tremendously from the great feedback I have received at APS meetings when presenting posters and talks, the constructive and thoughtful comments of the peer review process when submitting articles in APS journals, the professional advancement opportunities which helped my academic career by enhancing my national and international presence, and most importantly, by the great scientists with whom I have established long lasting friendships and scientific collaborations. As stated in the Odyssey by Homer “Many were the men whose cities he saw and whose mind he learned.” I am still travelling but I can see Ithaca on the horizon.

During my career I have been interested in identifying the basic mechanisms by which various insults such as hyperoxia, pathogens (such as mycoplasmas, influenza, respiratory syncytial virus, SARS) damage the blood gas barrier resulting in acute lung injury, and Adult Respiratory Distress Syndrome [4–9]. Based on mechanistic insights, I have strived to develop or repurpose drugs to decrease the onset and progression of lung injury. Thus, in 2006 I responded to a request for proposals from the National Institutes of Health, to identify the basic mechanisms by which exposure of animals to halogens (such as bromine and chlorine) damage their cardiorespiratory systems and develop therapeutics which when administered post exposure decrease the severity of lung injury and death from respiratory failure [10]. In this brief review, I will summarize the results of current studies implicating free heme as a major mediator of acute lung injury and our efforts to develop recombinant forms of human hemopexin, the most effective scavenger of free heme, as a countermeasure.

Exposure to chlorine (Cl₂) represents a major threat to public health:

Global production of Cl₂ exceeds millions of tons per year. Uses for Cl₂ include pulp bleaching, waste sanitation, chemical and pharmaceutical manufacturing, and drinking and recreational water treatment. Between 1940 and 2007, the accidental release of Cl₂ in 30 large cities worldwide[11], and the deliberate release of Cl₂ during acts of terrorisms by insurgents in Iraq and Syria, caused significant mortality and morbidity to humans and animals [12–14]. Exposures to swimming pool water treatment and household accidents due to mixing bleach with acidic cleaners resulted in wheezing and exacerbate the severity of asthma and chronic obstructive pulmonary diseases[15]. The US Department of Homeland Security identified industrial sites dedicated to Cl₂ as possible targets of terrorist activity, with some projections estimating that a successful attack could result in thousands of fatalities and 100,000 hospitalizations. Current treatment for toxic Cl₂ gas exposure is supportive and targeted therapies are unavailable. For this reason, there is a concentrated effort to develop animal models to identify the basic mechanisms by which chlorine damages the cardiorespiratory systems and develop appropriate countermeasures.

To draw valid conclusions from animal studies that will prove relevant to humans, it is important to use exposure models that approximate conditions found during release of Cl₂ in the atmosphere. Using validated plume models [16] in the Graniteville accident it was determined that Cl₂ levels during a 30 minute exposure period were 4428, 550 and 161 ppm at 0.2, 0.5 and 1 km down-wind from the epicenter of the accident. Persons exposed to 550 ppm Cl₂ for 30 minute required hospitalization and a number of them developed ARDS and required tracheal intubation and mechanical ventilation[11]. Survivors are more prone to bacterial infections and the development of pulmonary fibrosis, systemic hypertension and decreased FEV1.0, indicative of COPD disease [16]. People that smoke and those with respiratory infections and asthma are much more sensitive to the toxic effects of Cl₂ and may experience airway hyperresponsiveness to irritants. Exposure to high levels of Cl₂ gas led to vascular injury and depression of cardiac function [17].

Animal Models of Cl₂ Toxicity.

Mice, rats, rabbits, pigs, sheep, dogs, and monkeys have been used to study the cardiopulmonary sequelae of exposure to Cl₂ and their response to various treatments (reviewed in [13;18]). Experiments in rodents show that the degree of lung damage after Cl₂ exposure depends on total dose (i.e., concentration x time of exposure) as well as the specifics of exposure concentration and time [19]. Acute exposure to doses of 1ppm may cause irritation [20] and chronic exposure to even lower doses (0.1–0.4ppm) may lead to ocular irritation and degeneration of the airway epithelium. Irritant effects are primarily mediated through transient receptor potential ankyrin 1 (TRPA1) ion channels and their interplay with neurons within the airways that mediate the noxious reflex to inhaled irritant exposure, including the cough reflex[21]. Rodents exposed to 500–600 ppm Cl₂ for 30 minutes survive the exposure although they develop bradypnea and signs of respiratory distress [22;23]. At various times in the first 48 hours post exposure, rodents develop airway epithelial sloughing, interstitial and alveolar edema secondary to injury of the microvascular and alveolar epithelial cells, massive infiltration of inflammatory cells (mainly neutrophils), hypoxemia and respiratory acidosis, impaired surfactant function, decreased active Na⁺ transport across alveolar epithelial cells, airway hyperreactivity, alveolar hyper-coagulation and systemic hypo coagulation [24–28], resembling the symptoms of ARDS.

More than 50% of mice or rats exposed to 600 ppm Cl₂ for 30–45 minutes die from respiratory failure within 5 days post exposure. Animals that survive the initial Cl₂ insult exhibit resolution of many of the initial symptoms observed in the acute phase of injury post exposure, but often develop recurring, or progressively worsening elevated airway resistance in response to methacholine challenge, mucous metaplasia, abnormal epithelial repair, airway fibrosis, bronchiolitis obliterans and demonstrate changes in gene expression [26;29–32].

Mechanisms of Cl₂ toxicity.

Similar to typical ARDS, Cl₂ gas-induced lung injury is associated with a profound oxidant stress [33–35] detectable in rodents, human subjects, and human cells during and post exposure to Cl_2. Lung Injury is initiated by Cl₂, hypochlorous acid (HOCl), and hydrochloric acid formed by the reaction of Cl₂ with water in the epithelial lining fluid[36–39]. HOCl is also produced enzymatically by myeloperoxidase in resident or inflammatory cells (mainly neutrophils) recruited to the lung by the inflammatory response. Cl₂, HOCl_, accumulate in close proximity to the apical membranes of airways and alveolar epithelial cells where they provoke a robust inflammatory response resulting in inactivation of pulmonary surfactant[26;40], damage to lung ion channels [35;41], increased permeability to plasma proteins, and inability to clear alveolar edema fluid[41]. However, they are short lived and not capable of crossing the air-blood gas barrier. Thus, other agents must be responsible for the propagation of lung and systemic injury post exposure.

Reactions of HOCl with sulfhydryl groups, free amino groups of proteins, amino acids as well as plasmalogens lead to formation of long-lasting pro-inflammatory molecules, such as chloramines, chlorinated fatty acids, and aldehydes[42]. Chlorinated fatty acids have been detected in the plasma of mice, rats, pigs, and humans post exposure to Cl₂ and have been shown to increase fragility of red blood cells by oxidizing spectrin leading to the release of free hemoglobin in the plasma which is rapidly oxidized to heme [35]. (Figure 1).

Figure 1. Plasma cell-free heme and chlorinated lipids are elevated in humans and mice exposed to Cl₂ gas.

Blood was collected from patients exposed to Cl₂ gas 3–4 h post exposure, stored for 72 h at 4 °C at which time it was analyzed. Blood from human volunteers as controls was treated in the same fashion. (A). Cl₂ exposed individuals had elevated plasma levels of 16Cl-FA (chloro-palmitic acid; n = 4–5) and 18Cl-FA (chloro-stearic acid; n = 4–5). (B) Plasma free heme levels in persons exposed to Cl₂ were higher than the age and sex-matched human controls (n = 5–6) (Modified from Fig. 1 of Aggarwal et al. [35]). Individual values and means± 1 SEM. (C) Adult C57BL/6 mice exposed to Cl₂ gas (400 ppm, 30min) and returned to room air. At the indicated times, total (free and esterified) 16Cl-FA and 18Cl-FA were measured (n=3–9 mice for each time point). Modified from Figure 3 from Ford et al. [42]. Values are means ± 1 SEM, (D). C57BL/6 mice were exposed to air or Cl₂ (500 ppm for 30 min) and returned to room air. Blood samples were drawn at 1, 6 or 24 h post exposure and plasma heme was measured; n=6–8 for each time point. * Significantly different from the air value (p<0.01) by one way analysis of variance (ANOVA) followed by the Dunn–Šidák test for multiple comparisons. Values are means ± 1 SEM. Modified from [67]. (E). Adult C57BL/6 mice were exposed to Br₂ (600 ppm, 30min) or Cl₂ (400 ppm, 30min) and returned them to room air. Mice were sacrificed 24 h post exposure, RBCs were isolated and their mechanical fragility was measured following agitation. Individual values and means± 1 SEM Modified from Aggarwal et al. [35]. (F). RBCs were isolated from adult C57BL/6 mice and incubated ex vivo with either 16FA (palmitate) or 16Cl-FA (chloro-palmitate) (1 μM each). After 4 h incubation mechanical fragility was measured. Modified from [35]. Individual values and means ± SEM. The results were analyzed by one-way ANOVA followed by Tukey post hoc testing.

Role of Heme in the propagation of lung injury.

Heme is an iron containing molecular component of hemoglobin and plays an essential role in the transport of oxygen, when encapsulated in red blood cells (RBC). However, a number of pathological conditions increase RBC fragility resulting in the release of hemoglobin in the plasma which is rapidly oxidized to heme. Non-encapsulated heme is a pro-oxidant, capable of generating reactive species and iron, resulting in lipid peroxidation, protein and DNA injury [43;44]. It also acts as a feed-forward agent, further damaging RBCs leading to the release of additional hemoglobin and heme in the plasma, thus accentuating pulmonary and systemic injury and mortality. Non-encapsulated heme contributes to the onset and progression of lung injury after lipopolysaccharide exposure [45], Libby amphibole asbestos exposure[46], hyperoxia-induced lung injury[47], acute chest syndrome and pulmonary hypertension in sickle cell disease[48;49], trauma-hemorrhage[50], bacterial infection[51], neuropathic pain in HIV[52], and a number of other conditions. In addition, free heme has been detected in the plasma and bronchoalveolar lavage fluid of patients with ARDS[53;54], sepsis[55], COPD[56] and higher concentrations have been correlated with worse disease outcome. Furthermore, HO-1^(−/−) mice, which cannot break down free heme, are more susceptible to halogen-induced injury [57;58].

Hemopexin reverses Cl₂-injury in mice.

Free heme is maintained at low levels by serum albumin, haptoglobin, and hemopexin[59], a plasma protein with the highest known binding affinity to free heme (K_d near 10⁻¹³ M). The heme–hemopexin complex is transported to liver cells where it binds to CD91/LRP1, is internalized via receptor-mediated endocytosis and metabolized to biliverdin, carbon monoxide and iron by heme oxygenases (HO). Low levels of hemopexin have been associated with higher mortality in ARDS and sepsis [55;60;61]. A single injection of human hemopexin (5 mg/kg BW; administered intramuscularly in mice at 1 hour post exposure to Cl₂) decreased RBC fragility, returned plasma heme to air levels, decreased acute injury at 24 hours post exposure and improved survival (Figure 2). Transgenic mice expressing the human sickle cell gene have significantly higher levels of free heme in plasma after exposure to Cl₂ and develop severe lung injury and have high mortality. A single intramuscular injection of hHPX post Cl₂ exposure for these mice, decreased their plasma heme concentrations, the degree of lung injury and improved survival [49]. Similarly, a single injection of human hemopexin (hHPX), mitigated the development of both acute and chronic COPD-type injury in mice exposed to the halogen bromine by scavenging free heme and reducing the development of pulmonary endoplasmic reticulum (ER) stress[56] (Figure 3). hHPX has also been shown to decrease cardiopulmonary and systemic injury in a number of hemolytic diseases [55].

Figure 2. Injection of hHPX decreased heme levels, RBC fragility, acute lung injury and improves survival post Cl₂.

(A) C57BL/6 mice were exposed to air or Cl₂ (500 ppm for 30 min; open circles) and returned to room air. One h post exposure they were injected intramuscularly with human hemopexin (hHPX; 5 μg/g BW in 50 μ sterile saline). All mice were then sacrificed at 24 h post exposure, their lungs were lavaged and an arterial blood sample was drawn from the descending aorta. (A) RBC fragility (B) plasma K⁺ (C) plasma heme, (D) BALF proteins, (E) # of inflammatory cells in the BALF for the indicated groups. Each point represents results from a different mouse. Means ± 1 SEM; these data indicate that exposure to Cl₂ resulted in significant hemolysis and injury to the blood gas barrier at 24 h post exposure. Cl₂ exposed mice also developed compensated respiratory acidosis (F) and a 40% decrease of lung intact mitochondrial DNA (G). Individual values and means + 1 SEM. A single injection of hHPX (5 μg/ g BW) at 1 h post exposure had a beneficial effect on all of these variables and (H) improved survival considerably: at 15 days post exposure less than 40% of mice injected with saline were alive vs. 80% of those receiving hHPX. From Matalon et al. [67].

Figure 3. Injection of human hemopexin decreases chronic lung injury and improves survival in bromine (Br₂) exposed mice.

Male C57BL/6 mice were exposed to air or Br₂ gas (400 ppm, 30 minutes) and then returned to room air. Some Br₂-exposed mice were given an intraperitoneal injection of purified human hemopexin (Hx) (4 μg/g BW) 1 hour or 5 days after Br₂ exposure. All air-exposed mice and some Br₂-exposed mice received saline injection as an appropriate control. Fourteen days after Br₂ exposure, (A) mouse BALF protein levels and (B) total cell count were elevated in saline-treated mice but were significantly lower in Hx-treated mice. (C) Hx-treated mice had decreased lung deposition of collagen on Masson’s trichrome staining (n = 5–8) and (D) lower lung hydroxyproline levels compared with saline-treated mice, 14 days after Br₂ exposure. (F) Lung pressure-volume (P-V) curves, measured by flexiVent, demonstrated that Br₂ exposure increased lung volumes and lung compliance in the saline-treated mice but not in the Hx-treated mice (n = 5–10). (C) Hematoxylin and eosin (H&E) staining of lungs showed that Hx prevented Br₂-induced alveolar septa damage (n = 5–8) and (E) reduced alveolar Lm (n = 5–8). Scale bars 100 μm. In addition, (G) Hx lowered plasma elastase levels (n = 14–16) and (H) BALF elastase activity (n = 10–13) in Br₂-exposed mice. The Kaplan-Meier curves (I) demonstrated that Hx reduced mortality after B_r2 exposure when injected either at 1 h or 5 d post exposure (n = 42 for Br₂ + saline; n = 20 for Br₂ + Hx [1 hour after]; n = 17 for Br2 + Hx [5 days after]). *P < 0.05 versus air + saline and †P < 0.05 versus Br2 + saline by 1-way ANOVA followed by Tukey’s post hoc test. PV curves were analyzed by 2-way ANOVA with Bonferroni’s post hoc test. Overall survival was analyzed by the Kaplan-Meier method. Differences in survival were tested for statistical significance by the log-rank test. (from Aggarwal et al. [56]).

Mechanisms by which hHPX reverses lung injury.

Hemopexin binds equimolar amounts of porphyrins and metalloproteins. It is somewhat surprising that the administration of 5 μg/g BW of human hemopexin (total of 125 μg in a 25g mouse, approximately 1μM concentration and 10 times lower than the endogenous hemopexin levels in mice) could scavenge about 15 μM of plasma heme and prevent or reverse acute and chronic lung injury post halogen exposure. Multiple reasons could explain these effects of hHPX: First, it is possible that the endogenous circulating hHPX is already bound to free heme released from basal physiological hemolysis (due to RBC ageing) and is not ready to scavenge any de novo acute hemolysis. Therefore, exogenous administration of hHPX has profound effect in limiting heme levels. Second, elevated plasma heme has been shown to intercalate and destabilize RBCs membrane contributing to ongoing hemolysis and providing a feed-forward mechanism for more heme release[62]. Therefore, hHPX would not only scavenge free heme but also stabilize RBC membrane preventing further hemolysis. Third, administration of hHPX stimulates the release of endogenous mouse hemopexin by unknown mechanisms [56]. Finally, it is entirely possible that the lung protective effects of hHPX may not entirely depend upon heme scavenging and may also involve other anti-inflammatory pathways. Administration of 1μg-5μg/g BW of hHPX has completely abrogated lung injury by suppressing Toll-like receptor 4 (TLR4) and Nuclear Factor kappa B ( NF-κβ) [51;63;64].

To further investigate the latter, we performed robust global-discovery proteomics in lung tissues to identify specific targets of Cl₂ and the extent to which they are modified by human hemopexin The unbiased global proteomics analysis of lung tissues showed that exposure of mice to Cl₂ resulted in the modification of 237 lung proteins at 24 hours post exposure. Ninety-two of these proteins were increased and 145 were decreased in abundance at 24 hours post Cl₂ as compared to air. Eighty-three of the 237 lung proteins that were significantly changed (increased or decreased) in the lungs of mice at 24 hours post Cl₂ were modified following a single injection of hHPX at 1 hour post exposure. The top 40 most significantly changed proteins are shown in the two dimensional (2D) hierarchical heat map (HCA-HM, Figure 4A) and PCA analysis (Figure 4B). These data show clear differences in the lung proteome of mice exposed to Cl₂ and then returned to room air for 24 hours vs. those treated with hHPX. hHPX attenuated but did not completely reverse or prevented the effects of Cl₂ on the lung proteome. The top pathways modified by hHPX in the proteome of Cl₂ exposed mice are shown in the two 2D bubble diagram in Figure 5A. This analysis suggests that hHPX treatment modified a number of important processes associated with eIF2, integrin signaling and mitochondrial function, but did not completely reverse or prevent the Cl₂ effects on these biological variables. Western blotting studies using antibodies against elF2a and phosphorylated elF2a showed that at 24 hours post exposure to Cl₂ there was a large increase of the phosphorylated form of elF2a in the lungs of vehicle-treated mice, but not in the lungs of mice treated with hHPX (Figure 5B). An increase of phosphorylation indicates activation of elF2a by one of the four kinases and results in an overall decrease of protein transcription and translation [65]. These findings indicate that the beneficial effects of human hemopexin are complex and extend beyond its ability to scavenge heme.

Figure 4. PCA and heat map post chlorine (Cl₂) and human hemopexin (hHPX).

Male and female C57BL/6 mice were exposed to air or Cl₂ (500 ppm for 30 min) and returned to room air. One hour later, they received a single dose of human hemopexin (5 μg/g BW in 50 μL saline) or saline intramuscularly. Twenty-four hours later the mice were euthanized and their lungs were removed and proteins were processed for global proteomics analysis as discussed in the materials and methods. A: the two-dimensional hierarchical analysis heat map demonstrates which proteins are increased (red) or decreased (blue) in 24 h post Cl₂- versus air and hHPX-24 h post Cl₂ versus veh-24h post Cl2 (q < 0.1, P < 0.05, top 46 proteins). Each column represents a different mouse. B: the PCA complements the heat map by using a similar cluster approach that determines which animal (based on protein quantification for all proteins in the top list) is similar across all animals analyzed. Notice tight clustering for air, 24 h post Cl₂, and hHPX-24 h post Cl₂ mice with a clear separation between the three groups. Each indicates a different mouse. From Matalon et al. [67].

Figure 5. A. Two-dimensional (2-D) bubble chart of top pathways attenuated/reversed by human hemopexin (hHPX) post chlorine (Cl²) exposure.

This 2-D bubble chart illustrates pathway “categories” (y-axis) versus canonical pathways (x-axis). Statistical significance is indicated by negative log10 of the right-tailed Fisher’s exact test’s P value associated with the proteins that were found to be either attenuated or reversed in mouse lungs following at 24 h post Cl₂ exposure following injection of hHPX. The arrows are pointing toward “key” pathways found to be associated with each pathway category. The orange and blue color codes are highlighting the predicted pathway activities as either increased or decreased, respectively, which is indicated by their calculated z-scores. For all categories z = 1. The white and gray color codes are associated with either an unclear or balanced mix of activities based on the proteins identified, or simply no predicted alteration in pathway activities, which does not take away from the significant associations between the proteins of interest and their respective pathways, which is left to be determined via validation regardless. (B). Quantitation of the phosphor-eIF2α/total eIF2α in the lungs of mice at 24 post exposure to Cl2. Individual points and means ± 1 SE; statistical analysis by one-way analysis of variance followed by the Dunn–Šidák test for multiple comparisons. From Matalon et al. [67].

Need for the development of recombinant hemopexin.

hHPX is isolated from human plasma. Thus, it is costly and not available in quantities to conduct studies in larger animals and Phase I clinical trials. Thus, we established a research agreement and active collaboration with AntoXa, a Canadian biotechnology company, to generate properly glycosylated recombinant human hemopexin in tobacco plants by PlantForm under exclusive agreement with AntoXa. Plant-based production systems for biopharmaceutical proteins are attractive alternatives to mammalian cell, yeast, or bacterial systems[66]. Benefits include low cost, scalability, absence of non-animal components, stability, and rapid deployment of new therapies. In addition, products will be manufactured in a form that is stable over long periods suitable for administration to exposed individuals in the field. So far we have generated two forms of recombinant hemopexin which have been shown to reduce acute lung injury when administered in mice post chlorine exposure [67]. However, they lack proper glycosylation and sialylation and they are cleared rapidly from the plasma and have decreased efficacy as compared to hHPX. Additional products are currently in development and this project will keep us busy for a long time.