Chapter 1: Critical Issues in Biology and Medicine

We live in a material world with various substances and forms. There are observable structures that underlie the superficial phenomenon of human life. Our five senses feel, perceive, and recognize the physical environment. The type of foods, style of clothes, design of tools, and organization of architectures represent their symbolic meaning, interrelation, and functional purposes in the cultural context. We iteratively learn and understand their abstract meaning through social interactions and activities via languages and tongues. Depending on the type of economic system, societies have different legal frameworks, currencies, and patterns of infrastructures in which people exchange necessary resources and energies for living. Such a complicated network of persons and objects forms many social structures and organizations. These bring people to a consensus, educate our next generation, and as a result, make up a region, nation, or state.

Similarly, living organisms in wild nature have self-assembled structures and orders at different scales. Macroscopic morphologies are commonly used as differentiable features for taxonomy, an inductive approach to categorize similar species and understand their physiological function. For example, the vertebrate is featured by the axial skeleton of the spine that hosts the neural trunk and defines the symmetric axis of development. Those vertebral columns provide necessary support to the biomechanics of vertebrates. With the invention of the microscope in the late 16th century, scientists discovered a whole new world of microscopic organisms called cells. Unicellular organisms like bacteria can harvest nutrition and energy on their own. They communicate with one another and respond to environmental conditions through biochemical signaling. Just like a human society, cells can also live together under prerequisite scaffolds. Following mechanical and molecular cues, stem cells differentiate into specialized lineages in different regions of embryos. Then, they fulfill their function and cooperate in a niche environment. Ultimately, they commit their fates to develop multicellular organisms. These patterns of integration and behaviors of cells resemble how human society is forged. After 200 years of exploration and investigation under microscopes, biologists concluded that the cell is the functional unit and building block of life. Under guided development, multiple lineages of cells could be self-organized into functional tissues and organs. Shaped by the evolution process of adaption and survival, their histology and anatomy more or less reflect the necessary functions to cope with environmental challenges. Therefore, the dynamics of microscopic morphologies and macroscopic traits of organisms are the annotations of evolutional history and the functional support of live phenomena. The disorder of morphologies and cellular compositions may indicate the imbalance of physiology and the presence of illness.

Just like a cultural heritage in society, the traits of living things could be inherited and transmitted from parents to their offspring. People used to believe that this inheritance of characteristics is a continuous blending of parental characteristics. But in that case, all the unique phenotypes of organisms will be averaged over generations and lose their competing advantages under selective pressure. In 1865, an Austrian monk, Gregor Mendel, concluded his hybridization work on pea plants and discovered that heredity results from discrete units of inheritance called an allele, a pair of genes (Gregor, 1941). Each parent passed one allele gene to their offspring by chance, and the combination determines the phenotype. This is called Mendel's law of segregation. The inheritance of one gene will not affect the inheritance of others. This law of independent assortment indicates that the inheritance of different phenotypes is an independent event in statistics. With the discretization of inheritance, traits could be preserved long enough over the selection process, thus maintaining the survival strategies and related structures. Mendel's binary coding of inheritance can be applied to almost all organisms. But this hereditary process must be carried out by physical substances. What the composition is, where they are located, and how it works remained an enigma to be unraveled at that time. The related phenomena must take place at a scale that cannot be seen by naked eyes. Since the cell is the functional unit of life, it is reasonable to infer that the recipe of inheritance could be kept within the cells.

In studying the proliferation process of cell division, cytologists noticed that a dark-staining thread appeared in the nuclei of eukaryotic cells. They termed it chromosomes, which means colorful stuff. The chromosomes universally exist in cells of plants and animals. Different species have different numbers of chromosomes. In the mid-19th century, with the advance of apochromatic lens and immersion objectives (water/glycerin/oil), a bright-field microscope could achieve much higher resolution and acuity of images. Combined with a novel staining dye of coal tar by-product, Walther Flemming could time-course observe chromosomes' fine structure dynamics in detail. He identified chromosome duplication and split in the mitosis process (Flemming, 1882). Daughter cells inherited an equal share of duplicated chromosomes in pairs. The appearance of the chromosome serves as a phenotypic hallmark of cell division. Following this approach, Theodor Boveri found that the chromosome is necessary for embryo development (Boveri, 1902). Walter Sutton asserted that maternal and paternal chromosomes would reduce by half in their gametes. After fertilization, the offspring received each unpaired chromosome to form paired ones (Sutton, 1902). This hypothesis represented Mendel's segregation law and was later verified by the experiments on the recessive mutation of X-chromosome determining the eye color of Drosophila (Morgan, 1910). These serial discoveries indicate that chromosome in cells is the physical material that carries the heredity substance and genetic information, instructing the representation of traits. The chromosome theory of inheritance, thus, provides the mechanism foundation of Mendelian genetics.

In search of the philosopher's stone, Hennig Brand, in 1669, serendipitously discovered the chemical element phosphorus from biological material urine. He called it Icy Nocta Luca because phosphorus glows a vivid green light in the dark through the chemiluminescence process. Later, Robert Boyle found that, by mixing phosphorus with sulfur in a paper, a fire could be ignited on-demand by rubbing. This is the very original prototype of matches. Since then, alchemists have adopted scientific approaches to isolate or purify elements from minerals or organic materials. They have invented symbols to denote the reactants and products, kept down the protocols, and used the up-to-date scientific apparatus to analyze their properties. By the time John Dalton proposed the atomic theory, around 40 chemical elements had been discovered already. Chemists believe that all the matter in the world is composed of chemical elements with specific physical and chemical properties. Atomic elements cannot be broken down into other substances. For organisms like humans, the top abundant compositions of atoms are oxygen (O), carbon (C), hydrogen (H), nitrogen (N), calcium (Ca), phosphorus (P), and sulfur (S). They compose water, proteins, carbohydrate, fat, and bones. There are ions critical for the maintenance of physiology, like potassium (K), sodium (Na), chlorine (Cl), and magnesium (Mg). Trace amount elements like iron (Fe), cobalt (Co), zinc (Zn), iodine (I), and selenium (Se) commonly serve as cofactors participating in the enzymatic oxidation and reduction events. The assembling of the same or different atoms results in molecules, which is the fundamental unit of a pure substance. Most biological molecules are synthesized or degraded from multiple steps of chemical reactions. Even the chemical formula is the same, and different spatial arrangements of atoms may lead to totally different biochemical properties. For example, the fuel of cells, D-glucose (C₆H₁₂O₆), has two stereoisomers, planar α-D-glucose, and chair conformation β-D-glucose. Their difference is in the orientation of hydroxyl groups on carbon number 1 of the pyranose ring. The former can be polymerized into digestible starch and the latter into cellulose. Humans cannot digest cellulose because our enzymes cannot cleave the β-acetal linkage between the monomers of β-D-glucose.

Based on the available chemical elements and stoichiometry protocols, Gerardus Johannes Mulder and Jöns Jacob Berzelius started to investigate the chemical building blocks of major biological substances. At that time, albumins, fibrins, and gelatins were considered the essential component of animal substances (Holmes, 1963). Water-soluble albumins could be coagulated by heat and precipitated by acid. Fibrins could be extracted from muscle fibers. The gelatin could be obtained from the collagens and connective tissue of animals. Mulder chose to analyze water-soluble albumins, fibrins, and acid-precipitable caseins. They all shared a chemical core with a formula of C₄₀₀H₆₂₀N₁₀₀O₁₂₀P₁S₁ (Tanford and Reynolds, 2003). Materials from different species only differ in the composition of phosphorus and sulfur. They kept the same composition ratio among carbon, hydrogen, nitrogen, and oxygen, which is 40:62:10:12. Mulder called these definite-proportion molecules proteins. However, it is still hard to perceive the insight behind this formula from the composition of elements. In 1860, Robert Bunsen and Gustav Kirchhoff developed a powerful spectroscope equipped with a non-luminous flame to analyze chemical elements in a flame test. Different chemical elements have distinct sets of emission lines in the luminescence spectra. The vaporized elements will absorb sunlight at exactly the same spectral wavelength of emission lines. This atomic analysis method required only a few micrograms of samples. It greatly enhanced the sensitivity and specificity of element characterization. After many sleepless nights of measuring the luminescence and absorption fingerprints of known atoms, they finally discovered new chemical elements, rubidium (Rb) and cesium (Cs). Ernst Felix Hoppe-Seyler adopted this novel approach for his molecular analysis of pigments in red blood cells. He laked the blood and contained the pigment solution in a glass cuvette with a 1-cm sample thickness. Passing the sunlight through the solution, he found that the dispersed pigments have two characteristic absorption bands at green and yellow wavelengths (Perutz, 1995). The same absorption spectra could be found in the blood of different animal species. Treating the solution with acetic acid or strong alkali will make the bands disappear and split the materials into iron-containing hematin and proteins. At that time, some biochemists called the precipitable plasma proteins under a high concentration of neutral salt globulin. Therefore, Hoppe-Seyler abbreviately termed it hemoglobin (Natelson and Natelson, 1980). With further investigations, he found that oxygen has a binding affinity to hemoglobin. George Gabriel Stokes followed his experiment and found that the deoxygenation of hemoglobin will change the absorption spectra from a double band to a single diffusive band, a signature of venous blood (Perutz, 1995). This spectroscopy investigation on hemoglobin confirms its oxygen-carrying role in blood circulation. It links the protein molecules to physiological functions in organisms. Biochemistry, thus, emerged as an academic discipline to rationalize the live phenomenon with molecules. Photonic spectroscopy became a critical approach to capturing biomaterials' molecular essence, revealing their atomic structures and electronic interactions with light.

After the success of hemoglobin, Hoppe-Seyler and his student Friedrich Miescher tried to analyze the molecular content of nuclei, the place where chromosomes appeared. Using alkaline extraction and acidic precipitation, Miescher successfully extracted the nuclein and found a high elemental composition of phosphorus and nitrogen in it (Miescher, 1869). Albrecht Kossel further isolated nucleic acids, deoxyribonucleic acid (DNA), and ribonucleic acid (RNA), from nuclein, which means the nuclein is a complex of nucleic acids and proteins. In the late 19th century, on the phosphate-deoxyribose backbone, he further identified five nucleobases, adenine (A), cytosine (C), guanine (G), thymine (T), and uracil (U), as the constituent compounds of nucleic acids (Kossel, 1881). Similarly, the proteins were found composed of 20 different amino acids, and they connected into a peptide chain through the peptide bonds. Both nucleic acids and proteins (histone) locate in nuclei and have their own basis molecules to expand a larger space of molecules. The protein looks like having more letters and could compose more words or sentences to execute specific functions. In contrast, nucleic acids have limited potential for variability. Therefore, proteins were once considered as the reasonable candidates that carried the heredity information.

In the microscopy study of Mendelian genetics, the chromosome was identified as the substance that carried the heredity information. Since the chromosome was condensed from the nuclear substance, the biochemical answer of genetic materials, thus, narrows down to either proteins or nucleic acids. In 1928, Frederick Griffith conducted a famous bacterial transformation experiment showing that something in virulent Streptococcus pneumoniae could still transmit the virulence properties to a non-virulent strain even deactivated by heat (Griffith, 1928). With advanced protein precipitation and solvent extraction techniques, Oswald Avery and his colleagues confirmed that it is the DNA that allows non-virulent strains to inherit the function and phenotypes of virulent ones (Avery et al., 1944). To further ensure that DNA is the only substance in charge of genetic transformation, a negative proof that negates protein's role is necessary. With specific radioisotope labeling on DNA (³²P tag) and proteins (³⁵S tag), Hershey and Chase elegantly proved that it is the DNA that T2 bacteriophage injected into E. coli bacterium for replication, not proteins (Chase, 1952). This milestone experiment demonstrates how the molecular labeling technique traces the biodistribution of marker molecules in the biochemical process of cell biology.

We all know that the computer is operated by a software system composed of task-specific programs. These packages of programs, at the front end, have user-friendly input/output interfaces. At the back end, they control the function of hardware in the computer and process the information with sequential instructions. To mediate and coordinate human–machine interaction, programmers wrote the codes following a certain kind of programming language. In an editing environment of programs, the software engineer can translate designed workflows into a logical arrangement of codes. The programming language such as C++, Python, or Java will transcribe the scripts into low-level assembly language or machine codes, instructing the allocation of hardware resources like central processing unit (CPU), register, cache memory, random-access memory, and hard disk. For embedded applications, the structure of hardware or gearing of state machines can be edited by hardware descriptive language like Verilog. It forms a specialized functional unit for a specific task. This information handling architectures in computers originate from the innovation of the Jacquard weaving machine at the beginning of the 19th century (Hobsbawm, 1962). He designed a punched card as code to instruct the work of the weaving machine. This clever design distills the essence of iterative weaving jobs and produces expensive brocade massively. At the same time, the people in computing science also want to create a computing machine (Geselowitz, 2019), which is another tedious and labor-intensive work. They translated the punched card design to record data and input programmed instructions. Making the device work requires a machine language to transcribe readable words into strings of binary bits (0 and 1). Every character and symbol meaningful to human communication has their dedicated codons. Later on, the information processor evolved from bulky tabulating machines to densely integrated semiconducting transistors. The low/high voltage states, representing 0 and 1, are stored in electronic devices and magnetic hard disks. With the advance in lithography technology, the semiconductor industry has integrated more than 10 × 10⁹ transistors in a CPU and shrunk the bulky computer to portable and hand-held sizes. Nowadays, we can do regular paperwork and instant communication based on information digitization and global Internet architectures.

Now that the DNA has been confirmed as the carrier of genetic information, from the computer science perspective, there must be an underlying language system, and functional hardware instructs the production of proteins and other biochemicals necessary for cell biology. The logic of genetic codes and the signaling networks of proteins rationalize the life phenomenon from cells to organisms. By x-ray diffraction pattern of DNA crystals, Watson and Cricks confirmed the double-helix structures of DNA (Watson and Crick, 1953). Two complementary strands of nucleotides are held together by hydrogen bonds between G-C and A-T base pairs. This conjugated twist-ladder structure indicates that each strand can serve as a template to replicate the other strand (Meselson and Stahl, 1958). The corresponding protein machine of replication, DNA polymerase, was discovered by Arthur Kornberg in 1958 (Lehman et al., 1958). Since most of the proteins are synthesized in the cytoplasm, the genetic information on DNA needs to be first transcribed to messenger ribonucleic acid (mRNA) in nuclei (Brenner et al., 1961) and then delivered to the cytoplasm to complete the protein translation in ribosomes (Palade, 1955). In the transcription machine of RNA polymerase, the ATCG letters in DNA will be conjugately polymerized into single strand mRNA with UAGC correspondingly. According to the three-letter codon, these UAGC sequences of mRNA will recruit corresponding amino acids in the translation machine of the ribosome. After the formation of peptide bonds between neighboring amino acids, it will compile a peptide chain as the linearized proteins before folding. For example, AAA codons on mRNA will bind the UUU transporter RNA (tRNA) carrying the lysine amino acid. The AUG and UAA codons serve as the START and STOP signs for the peptide synthesis. Assisted by chaperone proteins, the linear peptide will be folded into its final conformation with a designated function (Saibil, 2013).

This information flow from DNA and RNA to proteins serves as the central dogma of molecular biology. The up- and downregulation of protein levels can be feedback-controlled by the action on the chromosome (epigenetic modification), DNA motif (transcription factor), RNA (interference RNA), and protein (ubiquitylation) (Riggs, 1975; Ciechanover et al., 1980; Johnson and McKnight, 1989; and Fire et al., 1998). The signaling of regulation can be triggered by ligand–receptor (membrane proteins) mediated phosphorylation. For differentiated cells with a specialized function, only particular sets of genes in DNA are transcribed. The DNA strands are wrapped around histone proteins. These blueprints are further folded into coils of chromosomes and tightly packaged in the three-dimensional space of nuclei. Just like Verilog language can edit the on/off layout of field-programmable gated array (FPGA) chips, the epigenetic methylation of DNA or acetylation of histones could repress or enhance the DNA transcription activities. When a part of the DNA is loosened from histones, that region of genetic codes can be unzipped for downstream transcription and translated. The expression pattern will define a specific lineage of cells through programmed epigenetic annotation.

When you visit clinics or hospitals, doctors check your medical records, listen to your chief complaint, perform a physical examination, and conduct diagnostic tests like blood routine, medical imaging, bacteria culture, or tissue biopsy. This differential diagnosis process helps doctors narrow down the problem space, adopt a suitable pathological model, confirm the potential cause, and guide the treatment. A disease may present characteristic morphological hallmarks in medical imaging and molecular markers in the blood test or in vitro diagnosis (IVD). Since different diseases could share the same hallmarks, doctors need to employ various inspection tools and chemical assays to collect differentiable evidence. In 1761, Italian pathologist Giovanni Battista Morgagni published his seminal work The Seats and Causes of Diseases as Investigated by Anatomy to explain how anatomical change facilitates the diagnosis of diseases. He set the foundation of pathology as a medical science studying the site and cause of disease. Later on, Marie-François-Xavier Bichat treated tissues as more fundamental elements in human physiological function. He recognized the disorder symptoms as a localized condition that began in specific tissues (Shah, 1994). With the advanced microscope and effective staining dyes, pathologists look into tissues' cellular composition and histology to define a disease's microscopic features. Particularly for chronic diseases, structure disorder could reveal the functional alteration of cells and tissues. Over 200 years of investigation, these cytometry and histopathology protocols have become regular laboratory medicine practices in hospitals.

Knowing that DNA is the genetic material that orchestrates the life phenomenon, many diseases may find their pathological origin according to the modification and expression of genes. But how does molecular disorder lead to the occurrence of diseases? The cell is an intricate molecular robot fueled by glucose or lipids. Its epigenetic annotations on chromosomes determine the transcription profiles and the expression levels of protein toolkits. Various lineages of cells in tissues have different blueprints of epigenetic annotation to execute their specific functions. Cells may exploit protein building blocks to develop unique organelles and necessary signaling pathways to fulfill their tasks in a micro-environment. Therefore, if the DNA somehow has uncontrolled transcription or translation, the cells may deviate from its regulated operation and lead to histopathological hallmarks in tissue morphology.

Taking angiogenesis as an example, when cells undergo hypoxia conditions, the intracellular oxygen-labile α subunit of Hypoxia Inducing Factor 1α (HIF-1α) will be free from ubiquitination (Semenza and Wang, 1992; and Groulx and Lee, 2002). Accumulated HIF-1α will bind to the constitutive HIF-1β subunit. The HIF-1 heterodimer will then recruit coactivators and transactivate the genes of vascular endothelial growth factors (VEGFs). The endothelial cells in nearby vessels can sense the VEGF's concentration gradient and sprout new vessels for oxygen supply (Gerhardt et al., 2003). Then, the recruited pericytes will wrap around tip cells and stabilize the vessel tubes (Nehls et al., 1992). This signal cascade coordinates multiple cells' interaction and promotes new vessels' maturation to cope with the shortage of oxygen. In tumor pathology, the hypoxic nodules also recruit vessels (Folkman, 1971). However, the profiles of tumor cytokines and overwhelmed VEGF signaling dysregulate the process of vessel maturation. Tumor hypoxia generates irregular shape and organization of vasculature. Decreased pericyte coverage results in leaky vessels and enhanced fluid pressure in the tumor micro-environment (Morikawa et al., 2002). These pathophysiological features of tumor angiogenesis impede the efficacy of drug delivery. Without active delivery, therapeutics can hardly reach the center of tumors. Further studies identify the Regulator of G-protein Signaling 5 (RGS5) as one of the master genes responsible for the occurrence of abnormal tumor vasculatures (Hamzah et al., 2008). People also found that the pro-inflammatory cytokine is necessary for pericytes to enable effective angiogenesis (Kang et al., 2019). The angiopoietin-2 (Ang-2) could induce pericyte detachment and endothelium cell destabilization (Keskin et al., 2015). This case demonstrates that genomic disorder in cells like cancer could emit derailed cytokine profiles to remodel the tissues' molecular and cellular composition. The molecular and cellular hallmarks may vary with the type of cancer and the tissues where it resides. Their expression or infiltration levels could dissect the stage of pathology and instruct treatment regimens.

Many molecular pathways tightly coordinate cells in the micro-environment to accomplish a biological task in situ. They dynamically maintain the physiological balance and longevity of organisms. However, in DNA repair or cell division, the DNA codes may not replicate with 100% fidelity. Even though there are quality check and error correction mechanisms, the DNA mutation or polymorphism will occasionally occur inside or outside the protein-coding regions. If genetic codes are destroyed or mutated at functional sites, corresponding proteins may reduce the activity or lose their functions. Single point error may modify the molecular milieu of cells and pose a risk to other mutations and dysfunction. Accumulated genomic aberrations may propagate to affect the regular operation of pathways, disrupt the homeostasis, and develop disease phenotypes in tissues. Mathematically, this genetic etiology of point mutation on DNA could be treated as a perturbation in a nonlinear dynamic system. Before mutation, cells stay in a dynamic steady state of physiology under the constraints posed by the molecular pathways. Mutational perturbation not only changes the expression level of a single protein but also alters the differential rate equation in related molecular pathways (Heinrich et al., 2002). If the change is too dramatic, a new equilibrium state may not exist, and the cell will die. If survived, the cells are strange attractors in the chaos system, carrying different functions and marker proteins (Nikolov et al., 2014). Therefore, in molecular medicine, disease-specific mutations may change the epigenetic annotation and switch cells to another global steady state that does not follow environmental education. Like stem cells, they have more plasticity and more functional tools to build up their micro-environment. Therefore, the dysregulated protein expression of the mutated cells will become pathology hallmarks in the differential diagnosis.

The DNA mutation could be an inherited germline mutation or an acquired somatic mutation. Some germline mutations are lethal to embryo development and cause the death of an organism prenatally. Such kind of lethal alleles is usually essential to growth or development. Some Mendelian disorders are not fatal to neonates and may appear later in the maturation stages (Dietz, 2010). Diseases like Marfan syndromes (FBN1 gene mutation) or Huntington diseases (HTT gene mutation) belong to this type. The cause-and-effect relationship of these congenital diseases is well-defined and apparent. But for loosely defined diseases like cancer or coronary artery disease (CAD), a single mutation or DNA sequence variation is just a sufficient cause, not a necessary cause. The disease onset also depends on other somatic mutations or chronic conditions. For example, BRCA1 and BRAC2 are high-penetrance breast cancer predisposition genes. Their mutations lead to gene instability and cause other gene mutations. The risk of breast cancer by age 70 may be as high as 87% for BRCA1 and 84% for BRCA2 mutation carriers (Turnbul and Rahman, 2008). But the BRAC1 mutation is just a sufficient cause of breast tumors. Other mutation sites like TP53 or PTEN could also lead to breast tumors at a lower susceptibility. Sometimes, the mutation may bring competitive advantages. The missense variant in PCSK9 reduces the LDL cholesterol level and protects people against CAD (Kathiresan and Srivastava, 2012). Mutation in the hemoglobin-beta gene changes the structure of red blood cells into sickle shapes. This deformation reduces malaria's susceptibility.

There are contributory gene modifications that are unnecessary and insufficient for disease occurrence. Even though they are not the root of the cause, they could either suppress or enhance diseases' severeness. In this way, this modification could serve as prognostic hallmarks. Moreover, there are many correlated markers of disease that are the downstream metabolomics of dysregulation. Their expression profiles have a strong correlation with the disease pathology. The subtle alteration in biological pathways can sensitively detect the aberrant process in the cells and tissues (Johnson et al., 2016).

When these markers are discovered and exploited for disease diagnosis, their accuracy will be evaluated by sensitivity and specificity. High-specificity hallmarks could faithfully rule out diseases if the level of makers is below the threshold. This indicates the probability of a negative test if the patient is healthy. High-sensitivity hallmarks are suitable for positive screening diagnosis. It indicates the likelihood of a positive test if the patient has the disease. A combination of markers may increase both sensitivity and specificity to achieve a better confirmation diagnosis. It is worth noting that pathology is a dynamic process. Markers at the late stage of diseases may not be the markers at their early stage. Sensitive hallmarks may not be useful for early diagnosis. It depends on when, where, and how genome-altered cells initiate or promote the pathology.

The tissue is a micro-society of cells in which epigenetic annotations program their predefined function and molecular toolkits. Different kinds of resident cells play their roles by expressing unique signaling ligands, membrane receptors, or metabolic enzymes. Through the secretion of cytokines, cells perform intercellular communication, educate neighboring cells, and even instruct the cells at distant organs. These multiple-cell clusters and the communicating pathways compose a niche engaging the specialized function of tissues. Without the support of niche cells, the orchestrating cells that determine tissue functions will fail to commit their tasks. The hematopoietic stem cells (HSCs) execute their function of hematopoiesis in the bone marrow and spleen. They majorly reside in the perivascular niche of sinusoids consisting of endothelial cells, CXCL12-abundant reticular (CAR) cells, and Nestin-positive mesenchymal stromal cells. Surrounding nerve fibers, Schwann cells, megakaryocytes, and osteoclasts also regulate HSCs through several mechanisms. Secreting factors in this niche environment can maintain the potency and differentiation capacity of HSCs (Crane et al., 2017). Not only do normal stem cells need a particular niche, but the tumor cells can also create their micro-environment to support their proliferation. The acidic tumor environment could modify the biology of macrophages and turn them into tumor-supportive activation (El-Kenawi et al., 2019). The tumor-associated macrophages (TAMs) could suppress the immune responses and help the tumor escape immune surveillance. The TAMs could also guide the metastasis of Mena^INV high-Mena11a low breast cancer cells (Pignatelli et al., 2014) and facilitate their entering into circulation. Cancer cells could hijack the PD-L1/PD-1 immune brake signaling to inhibit the T-cell's attack (Freeman et al., 2000). People even found that the starving pancreatic cancer cells could secrete nerve growth factors and instruct the innervation into the tumor micro-environment. The nerve cells can funnel the serine to save the nutrient-deprived tumors and support their further development (Banh et al., 2020).

These examples demonstrate that the niche is a location where supporting cells maintain the physiology of master cells in multicellular organisms. Usually, it is a nest for adult stem cells to replenish the progenitor cells required for organs. The environment cues of niche could even guide the differentiation of multipotent stem cells. Without the support of niche, many specialized cells, or statistically outlier cells, could not maintain their population. Like human society, the elite or aristocracy leads the organization or nation. They are outliers both in social status and in the authority they possess. Their ruling power relies on supporting administrators, police, and soldiers. The execution team or the government are the niches for the leaders. Together, the cooperation between niches and outliers determines the prosperity and decay of a society. With the aging of organisms, the niche may lose its vitality due to genetic alteration or wear and tear. The stem cell population will decline, leading to the degradation of tissue integrity and chronic diseases. Then, the micro-environment will undergo a metabolic shift and impair immune surveillance, favoring the emergence of other outliers, tumor cells.

Tumor cells are immortal populations that survive after several genetic mutations, which eliminate proliferation regulation and escape immune surveillance. The accumulated abnormalities in DNA let them proliferate without control. The fast-growing colony competes for resources with neighboring cells and deprives the nutrient in the tissues. When the nodule grows above a critical mass, typically with sizes above 1–2 mm³, the diffusion of oxygen and nutrient cannot reach the center of the tumor. The hypoxia condition will activate the recruitment of blood vessels and initiate tumor vascularization. In a harsh environment lacking resources, tumor cells turn their glucose metabolism to a glycolytic way and modify the cytokine milieu. Conventionally, people treat cancer with radiation and chemotherapy, killing the highly proliferative cells. But this type of treatment is a double-edged sword because it not only kills cancer cells but also harms the normal tissues, especially the routinely regenerative immune system. Besides, the treatment pressure will push the mutation and increase the heterogeneity of tumor genetics. Some outlier clones of tumor cells may become quiescent, acquire the capability of drug resistance, or transit into the mesenchymal lineage. This plasticity of tumors will find a way to relapse in the destroyed immune system. The lethal metastasis will finally disrupt the operation of organisms and put an end to life.

Research labs keep searching for various biological hallmarks of cancers that could serve as druggable targets to contain tumor growth. According to Hanahan and Weinberg's conclusion, a tumor's biological hallmarks include resisting cell death, sustaining proliferative signaling, inducing angiogenesis, evading growth suppressors, avoiding immune destruction, deregulating cellular energetics, creating genome instability, enabling replicative immortality, promoting inflammation, activating invasion, and metastasis (Hanahan and Weinberg, 2011). The tumor cells may acquire a part of these capabilities and build up their micro-environments. Many targeting therapies focus on blocking these surviving niches of tumors. For example, the anti-VEGF treatment could block abnormal angiogenesis and normalize the blood vessels for efficacious drug delivery (Hurwitz et al., 2004). Anti-PD-L1/anti-PD-1 immune checkpoint blockade therapy stops the immune deception signaling from cancer (Freeman et al., 2000). The treatment should aim at restoring the niche environment for normal cells and getting the organisms back to homeostasis.

Conventional chemotherapy and radiotherapy can quickly kill most of the tumor cells but also harm the normal ones. The selective pressure of treatment could induce resistant clones and recur afterward. The novel targeting therapy and nanomedicines can precisely deliver the drugs to the tumor micro-environment. It significantly reduces toxic damage to normal tissues (Muhamad et al., 2018). They are widely applied in the imaging diagnosis and treatment of malignant tumors. The targeting delivery helps to enrich drugs at the tumor sites and resolve toxicity to normal tissues (Sun et al., 2014). Researchers have developed various nanocarriers as an effective pharmaceutical formulation to decrease the loss of drugs and improve the biodistribution of agents in vivo (Yetisgin et al., 2020). Nowadays, nanomedicines have developed into a big family and play important roles in the fields of pharmaceutics and vaccines. With the targeting therapy of nanomedicines, the therapeutic effect can be further boosted (Bahrami et al., 2017).

Bacterial infections are still critical issues in public health. Infection-induced sepsis or multi-organ failure still challenges the medical system. In the 20th century, antibiotics like penicillin were discovered and widely used for antimicrobial treatment (Lobanovska and Pilla, 2017; and Hutchings et al., 2019). Peptidoglycan is an essential structural element in the cell wall for most bacteria. It can keep cell integrity and prevent macromolecules from penetrating into cells. By blocking the formation of peptide bridges, penicillin is able to prevent the formation of new peptidoglycan in bacteria. Then, the bacterial cells begin to lyse since it is difficult for them to resist osmotic stress. During 1940–1970, great efforts were made, and tens of millions of micro-organisms were screened. Many new antibiotics were found, including vancomycin, streptomycin, tetracycline, and erythromycin. The mortality of bacterial infection was significantly reduced. However, antibacterial therapy still faces enormous challenges because of the long-term use and the abuse of antibiotics, which induce the clones of multidrug-resistant bacteria. The rise of super bacteria is a result of evolution during a long-term antibiotic selection, because bacteria own remarkable genetic plasticity that allows them to evolve and then escape from a series of environmental threats. That is why scientists are always exploring new therapeutic mechanisms to solve this problem (Duval et al., 2019). Except for pharmacodynamic issues, the delivery of antibiotics into the desired locations in the human body is also critical for successful therapy. The antibiotics need to cross the biological barriers to infection sites. Depending on the infection sites, the very first barriers of delivery could be thick stratum corneum in the skin (topical antibacterial), the epithelial cells of the gastrointestinal tract (oral antibacterial), or the mucosa for the respiratory tract (pulmonary antibacterial) (Kamaruzzaman et al., 2017). After entering the tissues, the biofilms formed by bacterial communities are another barrier. These biofilms are inherently resistant to the penetration of antibiotics (Jamal et al., 2018). Finally, antibiotics must cross the membranes of infected cells through diffusion or endocytosis. Therefore, overcoming different biological barriers is a critical issue for the “chemotherapy” of bacterial infection.

Currently, chemotherapy is a standard option for treating malignant tumors. Unfortunately, chemotherapy never cures cancer thoroughly. Antitumor chemotherapies have different types of acting molecules, including antitumor antibiotics (Bleomycin, Doxorubicin), topoisomerase inhibitors (Etoposide, Topotecan), antimetabolites (6-mercaptopurine, 5-fluorouracil), and alkylating agents (cyclophosphamide, chlorambucil). They all lack selectivity to tumors, which would cause apparent side effects on normal tissues and organs. Furthermore, those drugs also have poor stability in the human body and, thus, would become less effective during blood circulation (Zhang et al., 2018). Henceforth, it is necessary to find new strategies to effectively enrich the therapeutics at tumor sites but distribute less at normal ones.

Targeting therapy is considered as one of the essential strategies for overcoming several crucial drawbacks of traditional medicine and treatments, such as high dosages needed, ineffective biodistribution, poor stability, and severe side effects. In antimicrobial treatment, various kinds of targeting approaches have been developed based on insights about the characteristics and properties of bacteria. The druggable targets include the surface structure, surface charge, and surface antigens of resistant bacteria, which are all different from normal tissues or cells. Both Gram-positive and Gram-negative bacteria can be targeting treated by controlling the composition ratio of surface ligands on drugs (Chen et al., 2018). In addition, designing drugs based on macrophage targeting is also a popular choice. A part of bacteria may survive and invade the immune system after being swallowed by macrophages, which protects them from antibiotics and cause the chronicity and relapse of bacterial infection (Teng et al., 2017). As a result, some antibiotics are not effective for bacteria in macrophages. In this situation, ligands such as mannosyl ligands, macrophage scavenger ligands, and fucosyl ligands can identify the receptors from the surface of macrophages (Chen et al., 2018). Therefore, preparing drugs based on ligand–receptor recognition provides a perfect way to solve this problem (Kelly et al., 2011). Besides, the application of antibodies is another effective way to realize targeting therapy for bacterial infection. The human body produces various antibodies, such as lgM and lgG, during bacterial infection (García-Gil et al., 2019). Some of them play significant roles in targeting drug delivery, which offers possibilities for the treatment of bacterial infection. For example, the gold nanoparticles can be modified with antibodies specific to Staphylococcus aureus, which own a perfect treatment effect on Staphylococcus aureus through the targeting effect (Millenbaugh et al., 2015). All these strategies were designed and implemented to overcome the delivery barriers.

As for cancer treatment, targeting therapy has two primary focuses. One develops drug delivery strategies to overcome the pharmacokinetic barriers. The other focuses on the tumor-specific metabolic pathways to achieve selective toxicity or synthetic lethality in pharmacodynamics. We need both passive and active targeting in drug delivery for better pharmacokinetics. The enhanced permeation and retention (EPR) effect of nanometer-sized medicines is a commonly used strategy for passive targeting. In the 1980s, many groups studied and validated the EPR effects of tumor biology (Morales-Cruz et al., 2019). When a solid tumor grows to a size more than 1–2 mm³, the hypoxia-induced growth factors VEGF secreted by cells could trigger the budding of new blood vessels from the nearby capillaries, which is known as angiogenesis (Burton and Libutti, 2009). The process induces a rapid proliferation of new, irregular blood vessels that present a discontinuous epithelium and lack the basal membrane of normal vascular structures. Depending on tumor location, micro-environment, and tumor type, the resulting fenestrations in the capillaries have various gap sizes from 100 to 800 nm (Morales-Cruz et al., 2019). The fenestrations of vessel walls could leak plasma nutrients to the tumor interstitium. This enhanced permeation allows larger-sized therapeutics to enter the tumor micro-environment. On the other hand, the lymphatic vessels in normal tissues constantly drain the extracellular fluid. The recycling of extravasated solutes and colloids can return to circulation by the lymphatic system. However, compared with normal tissues, the uptake of interstitial fluid in tumor sites is limited due to abnormal lymphatic function and blood vessel leakage (Aizik et al., 2019). Under these circumstances, the colloids with specific sizes would accumulate in tumor sites and not be cleared effectively. This lymphatic stagnation results in enhanced retention in the EPR effect.

Although the EPR effect has been well validated in small animal models, especially in rodents, the passive EPR effect is unlikely to present in all tumor types (Tee et al., 2019). Besides, the micro-environment of human tumors is drastically different from that of murine tumors, which will affect the optimal retention size of nanocarriers, biodistributions, and metabolic rates. For example, the human body has many physiological mechanisms to prevent and remove foreign materials in the body. In different organs of human body, drugs as foreign insults would be actively cleared by the reticuloendothelial system (Tang et al., 2019b). To solve this problem, loading drugs into nanocarriers modified with polyethylene glycol (PEG) is a major popular method to avoid uptake by the reticuloendothelial system. Besides, the EPR effects usually lead to high fluid pressure in the tumor micro-environment. The convection pressure will surpass the diffusion and prevent drugs from reaching the core of tumors (Subhan et al., 2021). Developing novel environmentally responsive nanomedicines integrating passive and active targeting effects has been regarded as a smart solution. After arriving at tumor sites through the EPR effect, the active targeting effect allows nanoparticles to bind the surface of tumor cells with the ligand–receptor affinity. Under specific stimulation, drugs are able to be released from nanoparticles. Taking the brain tumor sites as an example, the blood–brain barrier (BBB) formed by tightly bound endothelial cells prevents the penetration of therapeutics. This specialized vessel structure blocks a large number of molecules from entering the brain (Kievit and Zhang, 2011). The receptor-mediated transcytosis is a popular method to deliver medicine into regions with a brain tumor. The commonly used Transferrin receptor 1 (TfR 1) resides on the luminal side of the BBB. Modifying nanoparticles with transferrin can overcome the blood–brain barrier (Tang et al., 2019a).

According to extensive investigations, tumor cells usually have various overexpressed receptors based on the types of tumors, which provide excellent targeting sites to realize outstanding cell-specific targeting of drugs and reduce the off-target toxicity. Common molecular ligands of drugs for active tumor targeting include lectin, sugar residues, peptides, and proteins (Kumari et al., 2016; and Morales-Cruz et al., 2019). However, it is important to note that such active targeting would not improve overall accumulation around tumor tissues and could only enhance cellular uptake of the drugs following their passive extravasation through the EPR effect. Therefore, in most treatments for malignant tumors, both passive and active targeting strategies need to be synergistically considered.

Different targeting therapies have been explored and have made significant progress in recent years. The Food and Drug Administration (FDA) of the United States approved 50 nanomedicines for clinical use up to 2016 (Ventola, 2017). Targeting chemotherapy based on nanoparticles has greatly solved off-target toxicity problems caused by conventional therapies and brought antitumor research into a new stage. Nevertheless, some hurdles have yet to be overcome even under the assistance of passive and active targeting effects. Multidrug resistance is one of those problems in which tumors are capable of causing drug efflux through molecular pumps (Kunjachan et al., 2013) or a reduction of reactive oxygen species (ROS) damage through enhanced proteasome activity (Sang et al., 2021). These acquired functions of tumors counteract the pharmacodynamics of therapeutics and make them resistant to a broad spectrum of drugs.

When drug resistance happens, photoactivated therapy such as photodynamic or photothermal therapy (PDT or PTT) could be adopted to resensitize chemotherapy (Zhao et al., 2018). PDT requires a non-toxic photosensitizer, tissue oxygen, and light illumination to generate therapeutic ROS. During the process, photosensitizers in the ground state can transit to a singlet excited state under light irradiation. Some of the excited electrons may be coupled to excited triplet states and stay there for a long while. The long-lived (µs to ms) triplet-state electrons in photosensitizers can interact with oxygen and produce ROS to damage the targeting cells. As for PTT, it transforms light energy into heat and induces thermal ablation of tumor cells. Due to the short diffusion distance and life span, PDT can only react with targeted species in the adjacent area. In addition, most photosensitizers own poor stability or biocompatibility, which hinder their application in clinical practice. Phototherapy based on targeting enrichment can solve these problems to some extent. Nanocarriers could protect and deliver PDD or PDT agents to subcellular organelles with a high accumulation, through which the threshold energy of light activation can be greatly decreased.

Pharmacodynamic targeting therapy is realized through an understanding of the tumor-specific metabolism and surviving strategy. According to Hanahan and Weinberg's conclusion, tumors develop many survival strategies in their own micro-environment. Many metabolism-targeting strategies have been developed. For example, the growth of a tumor tends to depend on consuming a large amount of glucose, in which glucose will be transformed into lactate. Therefore, altering the glucose metabolism is an effective tool to realize a metabolism-targeting strategy. 2-Deoxyglucose is usually utilized for blocking glucose metabolism. Also, many pre-clinical studies have demonstrated that this targeting drug can inhibit tumor growth effectively (Luengo et al., 2017). In addition, altering enzyme expression in cancer cells is another fundamental approach. Take the isocitrate dehydrogenase as an example. It can be found in a large number of cancer cells. Developing drugs inhibiting isocitrate dehydrogenase provides new opportunities for cancer treatment. Dan Rohle and his co-workers (Rohle et al., 2013) developed a kind of isocitrate dehydrogenase inhibitor AGI-5198 that can suppress glioma growth.

As people learn more about signaling molecules in an immune system, immunotherapy has become a novel cancer treatment strategy that can promote and restore the ability of the immune system to recognize and eradicate tumor cells. In the late 19th century, William Coley first used bacterial toxins to treat cancer patients, which led to an antitumor immune response. Then, the cancer immune surveillance theory was proposed by Thomas and his co-workers, where the immune system can clear abnormal cells by recognizing tumor-associated antigens (Van den Bulk et al., 2018). Afterward, the role of T cells in the effective immune response was recognized, which brought about the clinical application of T-cell growth factor interleukin-2 (IL-2). IL-2 can induce T-cell proliferation, differentiation, and activation. It was the first cytokine successfully used in cancer treatment. Nevertheless, traditional immunotherapy based on IL-2 still has some problems such as low response rates and high toxicity. According to previous research, the process of interaction between tumor cells and the immune system includes three stages: elimination, equilibrium, and escape. In the first stage, abnormal cells would be recognized and eliminated by the immune system. This selective pressure urges the tumor cells to modify their gene expression and evolve to equilibrium. In the immune evasion stage, tumor-associated macrophages and immunosuppressive cells are recruited by tumor cells. By augmenting immune surveillance or blocking immune suppression, immunotherapy significantly improves the outcome of antitumor treatment. Following this concept, chemists have developed various modalities of immunotherapy to target the immune evasion mechanisms of tumors. Because of limited activation of the antitumor immune system, loss of targeted antigens, and inhibition of the immune reaction through tumors, antibody-based targeting immunotherapies are still not widely adopted. Several novel therapeutic methods, including cell-based targeting immunotherapy, have received much attention worldwide during past few decades. One of these methods is to augment the activation of T cells and revive their function against tumor cells. Then, the effector T-cells are transformed into memory ones and sustain the function of surveillance. This cell therapy solution relies on the engineering of chimeric antigen receptors (CARs) on allogenic or autologous T-cells. These equipped CAR domains on the surface of T-cells could bind to tumor-specific antigens and initiate adaptive immune responses to eradicate cancer cells (Larson and Maus, 2021). Another way of immunotherapy is to block the signaling of immune brake hijacked by tumor cells. The discovered immune checkpoint signaling targets include B7-CTLA4, PDL-1/PD-1, or FGL-1/LAG-3 ligand/receptor pairs (Mellman et al., 2011; and Wan et al., 2019). These passive strategies could avoid manufacturing complexities and expensive costs in CAR-T therapy.

Liposome was first investigated in the 1960s and then applied for drug delivery in the 1970s (Bernasconi et al., 2016). They are mainly prepared from biodegradable amphiphilic polymers and form membrane-like bilayer structures by self-assembly in an aqueous phase. The liposome structure has a hydrophilic inner cavity to load the hydrophilic drugs and the hydrophobic film to load hydrophobic therapeutics. The outer shells of liposomes are hydrophilic and can disperse in the water environment. Such a bilayer structure usually has a size range from 90 to 150 nm. Compared with other drug-delivery systems, liposomes have outstanding biocompatibility, no immunogenicity, high drug-loading efficiency, and the ability to co-encapsulate hydrophilic and hydrophobic agents. But the stability is still a considerable challenge to hinder the commercialization of liposomes. Some liposomes tending to aggregate, fusion, and drugs could also leak during the storage of liposomes.

Micelles are single-layered spherical colloidal particles prepared from amphiphilic copolymers through self-assembly. The hydrophobic inner cavity can load hydrophobic agents, while the hydrophilic shell provides stability and can avoid being cleared by the reticuloendothelial system. Due to the suitable size ranging from 10 to 100 nm, micelles can avoid rapid clearance by the reticuloendothelial system and prolong blood circulation. In addition, most antitumor drugs have poor solubility, which limits treatment efficacy. Fortunately, these micelles can improve the solubility of antitumor drugs after loading them into the inner cavity.

After arriving at the tumor sites, the loaded drugs must be released on-demand. Therefore, stimuli-responsive nanomedicines are designed and developed for drug delivery at effective dosages. These stimuli-responsive nanomedicines can prevent premature drug release and distinguish tumor cells from normal ones. The design requirements are based on substantial physiological differences between normal tissues and the tumor micro-environment. Human tumor tissues exhibit an acidic environment ranging from 5.7 to 7.8 owing to lactic acid accumulation caused by Warburg effects (Deirram et al., 2019). Poor lymphatic drainage and insufficient blood supply are the other reasons that induce the acidity of the tumor micro-environment. Two main strategies utilized for the design of pH-responsive nanomedicines are the (i) application of polymers with ionizable groups that could transform conformation and solubility under environmental pH changes; and (ii) design of drug-delivery systems with pH-sensitive bonds such as hydrazone, acetal, imine, and oxime bonds (Kocak et al., 2017). The carrier structures could be block copolymer (1–10 nm), star polymer, dendrimer (5–20 nm), hyperbranched, spherical micelle (10–100 nm), nanogel, vesicle micelle (50 nm–1 µm), and microgel (100 nm–50 µm). These pH-dependent drug release mechanisms are applicable in the tumor micro-environment and in subcellular lysosomes and gastrointestinal tracts.

Tumor-specific enzymes are another endogenous stimulus for the design of drug-delivery systems. For instance, the increased expression of glutathione tripeptide (GSH) is considered as a powerful reducing agent for the tumor to handle toxic radicals in a hypoxia environment. In tumor tissues, the concentration of intracellular GSH is at least four times higher than that of normal cells. Therefore, redox-responsive bonds, such as nitroimidazole groups, azo groups, and disulfide bonds, can be used to modify nanomedicines and control drug release. It should be noted that redox-responsive nanocarriers are usually utilized to release agents in the cell nucleus and cytosol, while pH-responsive nanomedicines tend to be applied for drug release in lysosome/endosomal compartments.

The controlled release of drugs can be triggered either by an endogenous stimulus or by an exogenous energy source. A magnetic field, due to its freely permeable nature, has been employed to heat magnetic-responsive nanocarriers. High frequency (50–1000 kHz) alternating magnetic fields could induce eddy-current heating, frictional heating, and hysteric-loss heating inside the iron oxide nanoparticles. Maghemite (Fe₂O₃) and magnetite (Fe₃O₄) are the main iron oxide materials for the preparation of magnetic nanoparticles. They have been approved by the FDA for their non-toxicity, stability, and ease of modification. Combined with heat-responsive polymers, magnetic hyperthermia can control the release of drugs (Mertz et al., 2017). In addition to the alternating magnetic fields, other non-invasive external stimuli, such as light or ultrasounds, have been successfully used to promote drug release from nanoparticulate systems (Wang and Kohane, 2017).

Molecular imaging aims to observe the dynamic molecular distribution and metabolism in vivo. With the help of contrast agents or molecular probes, pharmacologists can visualize and track the pharmacokinetics and pharmacodynamics of administered drugs. They can verify whether the nanomedicine works as designed, overcomes the barriers, accumulates at the targeting sites, and kills pathogens or cancers. Currently, the major imaging modalities include Positron Emission Tomography (PET), Single-Photon Emission Computed Tomography (SPECT), Magnetic Resonance Imaging (MRI), and optical microscopy. PET and SPECT use the gamma photons emitted from radio-labeled tracers to map the biodistribution of hallmark molecules. The detection limit of labeled compounds could be down to picomolar or nanomolar levels (Jiemy et al., 2018). Such high contrast sensitivity enables the in vivo imaging of metabolites, receptors, enzymes, and transporters. By designing tumor-targeting radiotracers, PET and SPECT can map the distribution of metastasis and follow up the treatment outcome. For PET, ¹¹C, ¹⁸F, and ⁶⁸Ga are the most commonly used radioisotopes as the molecular tag. They all have short half-lives, making them suitable for analyzing fast biological processes. During the decay of radioactive isotopes, positrons are produced and then annihilated when colliding with an electron. According to Einstein's law of mass-energy conservation, the disappearing mass of positrons and electrons will convert into two gamma photons. This pair of photons will travel in opposite directions to satisfy the momentum conservation. The difference in their arrival time on the surrounding PET detectors can reflect the location of tracers. Unlike PET, the tracers in SPECT emit single gamma photons, and a sensitive CCD camera can capture the omnidirectional illumination pattern. This technology typically involves rotating one or more camera heads around the patient, which helps reconstruct the 3D distribution of radioactivity in vivo (Wahl et al., 2011). Common radionuclides for SPECT include ¹²³I, ¹³¹I, ¹¹¹In, and ^99mTc. Generally, PET has a higher spatial resolution (∼1 mm), sensitivity, and accuracy in qualification than SPECT. The radionuclides of SPECT have longer half-lives than radiopharmaceuticals of PET, allowing SPECT to evaluate slow biological processes. Besides, SPECT scan is less expensive and is applied widely in molecular imaging. Overall, PET and SPECT have the highest sensitivity of molecular imaging at the cost of poorer structural resolution. Their functional information should be registered with x-ray computed tomography or MRI to obtain the precise structural location necessary for the diagnosis of diseases, evaluation of pathology, or confirmation of efficacy.

MRI applies strong magnetic fields (1–4 T) on organisms' bodies to prepare the resonant precession states of the spin magnetic moments in the atoms (Rodríguez-Galván et al. 2020). Incident microwave pulses at the resonant frequency can excite and synchronize the precession of nuclei, and the re-emitted electromagnetic waves can reveal the decoherence dynamics. By detecting longitudinal (T1), transverse (T2) relaxation time, water proton velocity, and density, MRI can differentiate the water micro-environment and reveal the structures of soft tissues. At ∼7 T magnetic fields, MRI can achieve a 100 µm in-plane spatial resolution. Furthermore, contrast agents such as gadolinium (III) and manganese (II) for MRI can improve the contrast quality of tissue images. Compared with x-ray CT, MRI could give more detailed structures in soft tissues. It also provides functional information about the neural connectivity in the brain.

Optical microscopy is an essential technology to observe cell and tissue morphology. Its submicrometer spatial resolution of images is not available for other molecular imaging modalities. In order to rationalize the physiological/pathological phenomenon and the treatment responses, cellular level evidence is necessary to prove the concept of the underlying mechanism. After the advancement of the fluorescence proteins and transgene labeling, optical microscopy can report genes' transcription and expression activities. The protein markers can further differentiate the identity and status of cells. This molecular information facilitates the analysis of cellular composition and molecular hallmarks in the micro-environment of tissues, which is critical for the investigation of tumor biology and stem cell niche. Unfortunately, the limited penetration depth is still the main disadvantage of optical microscopy. Under this circumstance, optical microscopy can only observe the cellular phenomenon in pre-clinical models like nematodes, Drosophila, zebrafish, mice, and isolated embryos and organoids. Rather than whole-body investigation, only specific sites can be analyzed when studying the mechanism of cell biology in micro-environmental changes of physiology and pathology. In the following chapters, we will explore how to use optical imaging and spectroscopy with nonlinear optical contrasts to increase the depth and dimension of optical molecular imaging. They expand the information dimension in the proof-of-concept pre-clinical investigations.