Portal de Periódicos da CAPES

Imagine physiology without imaging

2020; Wiley; Volume: 230; Issue: 3 Linguagem: Inglês

10.1111/apha.13549

1748-1716

Kathleen Cantow, Luis Hummel, Bert Flemming, Sonia Waiczies, Thoralf Niendorf, Erdmann Seeliger,

Climate Change and Health Impacts

The importance of diagnostic imaging is undisputable for those practicing medicine or those experiencing medical practice as patients or their families. Who could ever imagine modern clinical medicine without imaging techniques such as X-ray radiography, computed tomography, ultrasound and magnetic resonance imaging? Even microscopy is widely known as an indispensable tool for diagnostic histopathology. It appears, however, less commonly known that imaging techniques have been and continue to be pivotal for pioneering research in biomedical sciences including physiology. Yet, throughout the history of biomedical research, innovative (imaging) techniques may even have had a greater impact on scientific progress than new scientific hypotheses per se. A case in point in the early stages of biomedical research—or natural philosophy as it was called at this time—is light microscopy. Although still in its infancy, this technique enabled Robert Hooke (1635-1703) to discern 'cells' in a thin slice of cork and, later on, also in ferns and sundews. Reminded of the cells in a honeycomb he coined this term accordingly and published his discoveries in a book aptly titled "Micrographia: Or some Physiological Descriptions of Minute Bodies made by Magnifying Glasses with Observations and Inquiries thereupon." (London, 1665) [emphasis added]. At around the same time (1661), Marcello Malpighi (1628-1694) was the first who, via light microscopy, observed and correctly described capillaries—in a frog's lung! While the pulmonary circulation had been described much earlier by Ibn an-Nafis (Arabic: ابن النفيس , c.1210-1288), Realdo Colombo (1516-1559) and Michael Servetus (Spanish: Miguel Serveto, 1509-1553), Malpighi's discovery provided direct evidence for one pivotal aspect of the theory on blood circulation William Harvey (1578-1657) had developed some 30 years earlier. Further corroboration of Harvey's theory came from observations of blood flowing through capillaries by Antonie van Leeuwenhoek (1632-1723), also using the light microscope. Harvey's theory, published in his famous work "Exercitatio Anatomica de Motu Cordis et Sanguinis in Animalibus" (Frankfurt, 1628), was no less than a complete paradigm shift in physiology vis-à-vis contemporary theories and beliefs still adhering to the writings of Galen of Pergamon (Greek: Γαληνός, 129-c.216). Today van Leeuwenhoek is remembered mainly as the first who documented microscopic observations of microbes. Light microscopic detection of various bacterial species in the 19th century paved the way for Robert Koch (1843-1910). He proved that specific microbes are the cause of specific infections, earning him the 1905 Nobel Prize in Medicine and Physiology. Already then it was clear that infectious microbes must exist that are too small to be seen by the light microscope, but it took several decades before the electron microscope (EM) was invented to achieve the necessary spatial resolution. Helmut Ruska (1908-1973) was the first to develop EM applications for biological specimens, based on inventions by Ernst Ruska (1906-1988, 1986 Nobel Prize in Physics), Hans Busch (1884-1973), Max Knoll (1897-1969) and Reinhold Rüdenberg (1883-1961). In 1940, Helmut Ruska published the first comprehensive description of viral structures. Since then, EM—with its refinements and developments (cf. 1986 Nobel Prize in Physics, 2017 Nobel Prize in Chemistry)—has become an invaluable tool in biomedical research. For physiologists, disruptive developments in light microscopy including intravital, multiphoton and fluorescence microscopy (cf. 2014 Nobel Prize in Chemistry) have also opened new research avenues. The combination of super-resolved optical imaging at the level of single live cells, even single molecules, with single cell genomics including gene expression cartography,1 have become additional driving forces to probe (patho-)physiology and molecular function in real time. The early roots of non-invasive imaging and diagnostic radiology date back to Wilhelm Conrad Röntgen (1845-1923), who was the first to produce and detect X-rays or Röntgen rays in 1895. This invention ultimately also led to Godfrey Hounsfield's (1919-2004, 1979 Nobel Prize in Medicine and Physiology) development of computed tomography (CT) in 1973. Both X-ray radiography and CT continue to be cornerstones of clinical diagnostics. At the same time, various biomedical research fields have utilized and continue to utilize X-ray techniques. To name but one example, it was X-ray crystallography that allowed Rosalind Franklin (1920-1958) to obtain images of DNA that led to the famous description of the structure of the DNA and earned James Watson (b.1928), Francis Crick (1916-2004) and Maurice Wilkins (1916-2004) the 1962 Nobel Prize in Medicine and Physiology. Today, X-ray crystallography, often in combination with other modern techniques such as cryo-EM, is successfully employed to decipher structure-function relationships of much smaller biomolecules. It is somewhat ironic that Röntgen's discoveries were awarded with the (very first) Nobel Prize in Physics (the 1901 Prize in Medicine and Physiology went to Emil von Behring), even though the X-ray technology would have such a huge impact later on, both in clinical medicine as well as biomedical research. The next giant step in the history of non-invasive imaging took place about 80 years later, when Paul Lauterbur (1929-2007) developed a method to encode spatial information into a nuclear magnetic resonance (NMR) signal. NMR had already been known, and used to study biological specimens, for several decades. However, Lauterbur's new method was the beginning of magnetic resonance imaging (MRI). Among other scientists, Sir Peter Mansfield (1933-2017) pioneered advanced methods for MR image acquisition and processing (2003 Nobel Prize in Medicine and Physiology, jointly with Lauterbur). MRI scanners were first introduced into clinical medicine in the 1980s. Since then, MRI has experienced a plethora of technologic developments in hardware, imaging protocols and data processing, which have led to its widespread use in clinical medicine.2 Yet, MRI is, in current clinical practice, almost exclusively used to assess (pathological changes in) morphology only. MRI in (pre-)clinical research, on the other hand, makes use of an ever broadening range of techniques that enable assessments of (physiological and pathophysiological changes in) various functional parameters. Probably the most prominent of these techniques is the so-called blood oxygenation level–dependent (BOLD) MRI. This technique was introduced by Seiji Ogawa (b.1934) in 1990.2 BOLD MRI relies on the fact that deoxygenated haemoglobin (deoxyHb) is paramagnetic and, therefore, impacts on the effective transversal MR relaxation time T2*. T2* reflects the amount of deoxyHb per tissue volume, which can serve as a marker of blood oxygenation. Thus, BOLD MRI enables in vivo assessment of (changes in) blood oxygenation in real time; it is, furthermore, non-invasive, and does not employ ionizing radiation. Upon its invention, BOLD MRI was immediately used to study patterns of neuronal activity in the brain. This was not surprising: for the first time in history, patterns of neuronal activity could be visualized in conscious humans!2 The approach is based on the paradigm that an increased local neuronal activity triggers an increase in local blood flow, which in turn decreases oxygen extraction of the blood perfusing the region, thereby lowering local deoxyHb.3 This led to the broad field of functional MRI that today is widely used to study temporal and spatial neuronal activity patterns within the brain during a plethora of (patho-)physiologic conditions, tasks and paradigms. An increasing number of BOLD studies have recently focussed on mechanisms of cardiovascular control. Thus, several studies aimed at further elucidating characteristics of the arterial baroreflex. Aside from methodological questions pertaining to its sensitivity,4, 5 the renewed interest in this reflex is driven by new approaches to treat resistant hypertension by means of baroreceptor stimulation, selective vagal nerve stimulation and deep brain stimulation.6 Brainstem nuclei are known to govern the arterial baroreflex, but this has hitherto been almost impossible to explore in living humans. A recent study that made use of advanced BOLD MRI of the human brainstem paves the way to further elucidate, non-invasively, human cardiovascular control in health and disease.7 In another study in conscious humans, a dedicated BOLD technique demonstrated oscillations of cardiovascular parameters that are driven by central pacemaker activity in the brainstem.8 BOLD MRI continues to be a very valuable tool for preclinical studies that focus on pathophysiologies of cardiovascular control and potential new treatment options. Papers published in Acta Physiologica provide prime examples for the use of BOLD MRI in these research areas. For instance, BOLD measurements in spontaneously hypertensive rats showed that the lack of gut microbiome-derived butyrate influences the activity of cardio-regulatory brain regions; this led to the hypothesis that microbial butyrate may play a role in blood pressure regulation.9 In a rat model of ischemic stroke, dedicated functional MRI techniques were used in a longitudinal study to assess the benefit of early inhibition of the mitogen-activated protein kinase 1/2 administered at a clinically relevant time point.10 As a non-invasive technique, functional MRI complements various other 'classical' methods used to study cardiovascular control, ranging from—mostly invasive—in vivo to ex vivo techniques such as Ca++ imaging.9, 11, 12 Kidney MRI has taken a centre stage in biomedical research, making use of the unique opportunities provided by functional MRI, in particular, BOLD. This is motivated firstly by the pivotal role that renal tissue hypoperfusion and hypoxia are assumed to play in the pathophysiology of various kidney diseases and the constraints of 'classical' physiological methods in translational contexts.13-15 Several special features of renal haemodynamics and oxygenation make the kidneys exceptionally vulnerable to an imbalance between oxygen delivery and demand. They result—inter alia—in a major heterogeneity in tissue partial pressure of oxygen (pO2) among the renal layers (cortex, outer medulla and inner medulla) with very low pO2 in the medulla. An ever increasing number of preclinical studies—many of which published in Acta Physiologica—indicate that renal tissue hypoxia is an important early element in the pathophysiology of acute kidney injury (AKI), its possible progression to chronic kidney disease (CKD) and diabetic nephropathy.12, 14, 16-20 Many of these studies utilized invasive probes to measure renal haemodynamics and oxygenation in anesthetized animals. The probes typically include Clark-type electrodes or fluorescence optodes, ie gold standard methods for measurement of tissue pO2. As local tissue pO2 is quite heterogeneous even within a given renal layer, a significant drawback of these probes is that they obtain data within a rather small volume of tissue only.13 Here, BOLD MRI can play to its advantages: besides being non-invasive, it enables whole kidney coverage and high spatial resolution.15 Consequently, BOLD is increasingly used in preclinical studies that aim at elucidating renal (patho-)physiology as well as testing potential measures for prevention or therapy of renal disorders. For example, by means of BOLD it was demonstrated that specific pharmacological interventions targeting the cytochrome p450 (CYP)-eicosanoid pathway in renal ischaemia/reperfusion injury alleviates renal tissue hypoxia in rats.19 The translational importance of this result is underlined by findings in humans indicating that individual differences in CYP eicosanoid formation may contribute to the risk of developing AKI.19 In the majority of studies on animal models of renal diseases, BOLD data are complemented by other methods including, but not limited to, standard histology, EM, renal tissue-, blood- and urine-based biomarkers, and X-ray CT.12, 15, 18, 19 The second motivation to promote functional renal MR is the notorious lack of clinically available diagnostic tools that would allow early recognition of AKI and CKD, with sufficient sensitivity and specificity. In today's clinical practice, diagnosis of these disorders still relies on serum concentrations of creatinine. Although an ever increasing number of new blood- or urine-based biomarkers have been proposed, neither has hitherto advanced to provide point-of-care diagnosis for AKI. The early detection of AKI and, thus, the opportunity for the timely care for patients suffering from AKI, is widely recognized as a major unmet clinical need. Synergistic approaches that include functional renal MRI have been proposed to meet that need. Consequently, an increasing number of studies in humans evaluate the potential of renal MRI including BOLD as diagnostic tools for AKI and CKD.15, 21 Yet, to become clinically relevant as quantitative biomarker(s), functional renal MRI protocol(s) still require standardization with regard to physiological conditions and technical parameters. Last but not least, as a number of factors confound the renal T2* to tissue pO2 relationship, their quantitative impact in various (patho-)physiological scenarios must be determined.15 As the present short review illustrates, there can be no doubt that imaging techniques have had and continue to have a pivotal role in biomedical sciences including physiologic research. Images are generally highly impressing and convincing, and as the saying goes: seeing is believing. Yet, we should remain cautious to take results based on imaging (alone) at face value. Unintentional errors, inconsiderately overstated interpretations, and in some cases even fraud, are known to have occurred. An early and (in-)famous case in point is Ernst Haeckel's (1834-1919) depiction of embryos of various vertebrates in his book "Natürliche Schöpfungsgeschichte" (Berlin, 1868). He was justifiably accused to have 'embellished' these images in his attempt to support Charles Darwin's (1809-1882) theory of evolution in what is regarded as one of the most heated debates in (natural-)philosophy of all times. Copyright infringement could be considered an additional reproach for Haeckel's work in today's terms, even though copying images was not yet generally regarded as serious scientific misconduct in the 19th century. However his, at least marginally, fraudulent images are still misused today as foundation to disavow the theory of evolution.22 Another (in-)famous publication in 2009 reporting on a dead salmon's brain 'answering' when 'asked' to perform a social perspective taking according to an established paradigm appeared at first glance to merely ridicule BOLD MRI. On the contrary, this publication was a wake-up call for appropriate processing of BOLD data and its duly cautious interpretation.23 The necessity to careful interpret BOLD results was recently reiterated by a study in mice. Using 'classical' physiological methods to monitor cardiovascular parameters, the study demonstrated that a major portion of changes in brain BOLD data in an established somatosensory paradigm is induced by changes in arterial blood pressure rather than changes in neuronal activity.3 To conclude, the biomedical research community, and physiologists in particular, should be aware that promoting innovative imaging techniques across multiple scales in space and time is essential for the study of biological systems and disease. On the other hand, the community should also be aware that developing robust imaging tools and promoting the reproducibility and careful interpretation of imaging findings are equally essential. This includes the necessity of more critical evaluation of intriguing images as well as the conclusions derived from these images by means of other research methods especially those originating from 'classical' physiology. The findings obtained by this approach should be integrated with (big) data science, including artificial intelligence, predictive analysis and deep machine learning into a coherent picture of cells, tissues, organs and organisms for a better understanding of (patho-)physiology. With this exceptional level of innovation and rich opportunity for discovery, we cannot imagine physiology without imaging. The authors have no conflict of interest to declare.