Toward Therapeutic Pulmonary Alveolar Regeneration in Humans

Abstract

In humans, age results in loss of pulmonary alveoli; menopause accelerates loss of diffusing capacity, an index of alveolar surface area; and disease (e.g., chronic obstructive pulmonary disease) results in loss of alveoli. Thus, an important goal for investigators is to generate knowledge that allows induction of pulmonary alveolar regeneration in humans. Our enthusiasm for this goal and our assessment of its feasibility are based on work in several laboratories over the last decade that has disproved the notion that pulmonary alveoli are incapable of regeneration, and on the growing evidence that signals that regulate programs of alveolar turnover (loss and regeneration) are conserved from rodents to humans. We review animal models of alveolar loss and regeneration and their conservation during evolution, and hence their relevance to humans.

We believe an important long-term goal for investigators is to generate knowledge that allows induction of pulmonary alveolar regeneration, or that rescues failed alveologenesis, in humans. Our enthusiasm for this goal and our positive assessment of its feasibility are based on work in rodents in eight laboratories (1–6) (S. Rennard and S. Shapiro, personal communication) over the last decade that has falsified the notion that pulmonary alveoli are incapable of regeneration (7, 8), and on the growing evidence that signals that regulate programs of alveolar turnover (loss and regeneration) are conserved from rodents to humans (9–12).

CONDITIONS TARGETED

Age-related Alveolar Loss

In nonmutant animal species, whose thorax and lung volume do not increase during adulthood, an age-related loss of alveoli occurs, as indicated by an increase of distance between alveolar walls (Lm), or a decrease of alveolar surface-to-volume ratio, and most important, by a decline of alveolar surface area without a decline of lung volume (13–15). In rodents that continue to grow as adults, lung volume increases with age, as does Lm (14, 16, 17), but, more important, the volume of individual alveoli increases with age (18–20); these increases partly reflect loss of lung tissue elastic recoil with age (21) but are also due to the loss of alveolar walls, as shown by the decrease in the total number of alveoli with age (22, 23).

In humans without anatomic evidence of lung disease, morphometric studies clearly show alveoli get bigger, and alveolar surface area and the number of alveoli decline with age (22–24). Furthermore, the loss of alveolar surface area begins in the third to fourth decade of life and proceeds more rapidly in men than in woman (24); these anatomic findings, although based on few subjects, are supported by the early onset of loss of lung function and, until menopause, the more rapid loss in men than in women (25). Finally, loss of lung function in a general population is a strong predictor of mortality (26–31).

Menopause, which is age related, is an important cause of accelerated alveolar loss. Diffusing capacity, an indicator of gas-exchange surface area (32), diminishes with age in never-smokers. Men who have never smoked lose diffusing capacity at a rate of about 6%/decade; woman who have never smoked lose diffusing capacity at a rate of about 2%/decade before menopause and about 6%/decade postmenopause (25). These rates of loss of diffusing capacity parallel the age-related loss of maximum O₂ consumption (V̇max) (33, 34), although, for unclear reasons, the link between alveolar loss and the associated loss of V̇max is never made. That this menopause-related accelerated decline of diffusing capacity is due to a fall in the concentration of estrogen is supported by the observation that ovariectomy in adult mice results in alveolar loss and estrogen replacement results in alveolar regeneration (35); this observation also supports the notion that the effect of estrogen on lung function is evolutionarily conserved from mice to humans.

More recent work with women on the effect of estrogen on forced time expiratory flow rates adds additional support for evolutionary conservation of the effect of estrogen on alveolar architectural stability and regeneration. Thus, the forced time vital capacities reflect resistance to airflow in the conducting airways and lung tissue elastic recoil. Increased resistance to airflow is present in airways that are excessively narrow during expiration and can, in part, be due to the loss of the tethering effect of alveolar attachments caused by alveolar destruction (36–40). Elderly women receiving hormone replacement (estrogen plus progesterone) have a higher FEV₁ than similar-age women not receiving hormone replacement; this difference is not explained by lower rates of smoking or other health factors (41). Administration of estrogen plus progesterone (42, 43), estrogen alone (44), or an estrogen-like compound (44) to postmenopausal women increases their forced vital capacity and FEV₁. Even in women aged 24 to 35 yr, the use of an oral contraceptive containing estradiol and a progestin increased forced expiratory flow rates, especially flow rates at low lung volumes (45). The latter is especially important because expiratory airflow at low lung volumes reflects the patency of small conducting airways, which depend substantially on the tethering effect of alveolar attachments (36–40). Hence, these findings in young women point to the alveolar-maintaining effect, and perhaps alveolar-regenerating ability, of ovarian hormones. The loss of alveoli in mice after ovariectomy and their regeneration during estradiol replacement (35) suggest estrogen is the ovarian hormone responsible for maintaining alveolar structural stability, and for inducing alveolar regeneration, in women. This evidence that the estrogen-preserving effect on alveolar architectural stability and its alveolar-regenerating effect is evolutionarily conserved supports the usefulness of this mouse model and its relevance to women.

In addition to the role of ovarian hormones, in particular estrogen, on the alveoli of healthy women, low concentrations of ovarian hormones may play a role in the development and progression of chronic obstructive pulmonary disease (COPD) (46). Women constitute 75% of never-smokers older than 55 yr with clinical and lung function evidence of COPD (26, 47). That this observation does not reflect an age bias of smoking prevalence toward men is indicated by the standardized mortality rate, which is almost twice as high in women than in men, among individuals with COPD who participated in a survey in which the average age on entry was 56.6 yr (48). Thus, evidence is growing to indicate estrogen may delay the loss of, and improve, those lung functions that reflect maintenance of alveolar structure and, as a consequence, the number of alveolar attachments to small conducting airways. If proof-of-principle testing fails to falsify these conjectures, they could result in a decrease in suffering, death, and economic loss worldwide (49, 50).

Calorie-related Alveolar Loss and Regeneration: Evidence for Evolutionary Conservation from Mouse to Human

Calorie restriction in mice (51–53), rats (54–56), and hamsters (57) and starvation in humans (10–12) cause alveolar loss through, we believe, an endogenous program that is conserved from rodents to humans. Ad libitum access to food after calorie restriction–induced alveolar loss results in alveolar regeneration in rodents. To our knowledge, the effect on alveolar regeneration of ad libitum refeeding after starvation in humans has not been tested. However, the evidence that the calorie restriction–starvation endogenous program of alveolar loss, which offers survival advantages during commonly occurring periods of starvation, is evolutionarily conserved strongly supports the notion that its opposite, alveolar regeneration in humans after refeeding following starvation, will also be conserved—hence, the importance of understanding the gene expression responsible for alveolar regeneration in this model.

MODUS OPERANDI

Identification of Very Upstream Gene Expression That Is Specific and Determinative of Alveolar Regeneration

Here we present an approach we are using with the hope it will generate discussion, criticism, and better ways to move toward alveolar regeneration in humans. The use of global gene profiling, which should be a powerful tool to identify the initial gene expression that specifies alveolar regeneration, has, we believe, been hampered by the lack of knowledge of the time when, after a regenerative stimulus, the very upstream gene expression determinative of regeneration begins. This information would narrow the period that must be studied to identify, and understand, the regulation of genes that initiate alveolar regeneration. To identify this period, we have, as a first step, enumerated cellular processes that biological knowledge indicates are required for alveolar regeneration (i.e., transforming a flat segment of alveolar wall into an elongating fold [septum] that increases gas-exchange surface area). These processes include lung cell replication, angiogenesis, remodeling of the extracellular matrix, and guided cell motion. Then, we used microarray analysis in two mouse models of alveolar regeneration (estradiol-treated ovariectomized mice and previously calorie-restricted mice with ad libitum access to food) killed 3 h after the regenerative stimulus. These studies revealed the presence of gene expression in lung supportive of cell replication, angiogenesis, formation of extracellular matrix, and guided cell motion within 3 h of the onset of ad libitum access to food and estradiol treatment. We then used real-time polymerase chain reaction (RT-PCR) to assess expression of a subset of genes at earlier times after estradiol injection into ovariectomized mice. We identified gene expression supportive of cell replication within 1 h of estradiol treatment, and within 1 h of the onset of ad libitum access to food after calorie restriction. Three hours later, gene expression supportive of other processes required to form a septum was present.

We learned from this work with the ovariectomy and calorie-related models that we could go directly to using RT-PCR to identify the onset of changes in gene expression determinative of alveolar regeneration. Therefore, with our third model, alveolar regeneration produced by treating mice that have elastase-induced emphysema with all-trans-retinoic acid (5), we will study mice with elastase-induced emphysema 1 h after treatment with all-trans-retinoic acid.

Our modus operandi is as follows:

• 1. Provide a regenerative stimulus.

• 2. Use RT-PCR to determine the onset of signaling at the RNA level for cell replication, guided cell motion, extracellular matrix remodeling, and angiogenesis; our data indicate that, for each of the two animal models tested, signaling for cell replication is a very early response to a regenerative stimulus (e.g., within 1 h)

• 3. Establish the earliest time of appearance in each model, after injection of each regenerative reagent, of signaling at the RNA level for cellular processes supporting alveolar regeneration; this identifies the time of early gene expression determinative of alveolar regeneration

When this timing is established for all three models, or all reagents, we will

• 4. Perform microarray analysis over a short (1 h) period to get a global view of gene expression in lung in the models used

• 5. Use computational analysis of gene expression among the models or reagents

If we detect a common pattern of gene expression among two or more of the models or reagents, our proof-of-principle will be to

• 6. Manipulate the pattern of gene expression using small inhibitory ribonucleic acid and morphometrically determine the effect on alveolar regeneration

If we do not detect a similar pattern of gene expression among two or more of the models or reagents, it will mean the animal models relevant to human diseases must be studied individually, and candidate regenerative maneuvers must, as has so far been the case (1–7), be selected on available information and emerging insights.

ALVEOLAR REGENERATION IN RODENTS

Work over the past decade has falsified the notion that alveoli do not regenerate (1–6, 51, 52) (S. Rennard and S. Shapiro, personal communication, March 2006). During this period: (1) among seven laboratories, four reagents—all-trans-retinoic acid (Figure 1) (1–3) (S. Rennard and S. Shapiro, personal communication, March 2006), granulocyte colony–stimulating factor (58), adrenomedullin (4), and hepatocyte growth factor (5)—induced alveolar regeneration and abrogated key features of elastase- and cigarette smoke–induced emphysema in rodents; (2) two laboratories reported all-trans retinoic acid rescues failed alveologenesis in mice and rats (6, 59); and (3) one laboratory demonstrated estradiol induces alveolar regeneration after ovariectomy-induced alveolar loss in adult mice (35) and preserves alveolar formation after ovariectomy in rats at the time they are weaned (60). By contrast, studies in five laboratories have not found that all-trans-retinoic acid induces alveolar regeneration in mouse (61, 62), rat (63), guinea pig (64), or rabbit (65). The reason for this failure is unclear. Two of the studies, which were excellent, were performed at high altitude (Denver [61] and Albuquerque [63]). Because hypoxia inhibits alveologenesis (19, 36, 66, 67), it is possible this explains the failure of regeneration in Denver and Albuquerque. Our conclusion is that if nine miners strike gold, but five do not, gold is present. Finally, work in several laboratories, and the evidence that the regulation of alveolar loss and regeneration is conserved from mouse to humans, suggests we are not at the “beginning of the end,” but perhaps at “the end of the beginning” (Winston Churchill, 1942, wartime speech given at the Lord Major's luncheon) in the pursuit of alveolar regeneration in humans.

Figure 1.

Histologic sections showing (A) lung of a rat not treated with elastase; (B) lung of a rat in which elastase was instilled into the trachea and, 3 wk later was injected daily with oil, the vehicle for all-trans-retinoic acid, for 2 wk; (C) lung of a rat in which elastase was instilled into the trachea and that, 3 wk later, was treated daily with all-trans-retinoic acid for 2 wk. Lungs were fixed at a transpulmonary pressure of 20 cm H₂O. Reprinted by permission from Reference 1.