Raman spectroscopy is a minimally-invasive optical technique with great potential for in vivo cancer detection and disease diagnosis. However, there is no systematic study of the Raman spectra from different organs to date. We measured and characterized the Raman spectra eighteen naïve mouse organs in a broad frequency range of 700 to 3100 cm−1. The peaks of generic proteins and lipids appeared in Raman spectra of all organs. Some organs like bone, teeth, brain and lung had unique Raman peaks. The autofluorescence was strong in liver, spleen, heart, and kidney. These results suggest that organ specific Raman probe design and specific data processing strategies are required in order to get the most useful information.
© 2011 OSA
A Raman spectrum records the Raman scattering or the Raman effect which is due to the inelastic scattering of photons. The loss of photon energy which occurs during Raman scattering is transferred into greater vibrational or rotational energies of a molecule [1,2]. In 1922, Indian physicist Chandrasekhara Venkata Raman published his work on the “Molecular Diffraction of Light,” the first of a series of investigations with his collaborators which ultimately led to his discovery (on 28 February 1928) of the radiation effect which bears his name . The Raman effect was first reported by C. V. Raman and K. S. Krishnan, and independently by Grigory Landsberg and Leonid Mandelstam, in 1928. Raman received the Nobel Prize in 1930 for his work on the scattering of light.
Raman spectroscopy is commonly used in chemistry since vibrational information is specific to the chemical bonds and symmetry of molecules by which the molecule can be identified. In recent years, the application of Raman spectra in biomedicine and biology is increasing, which can be noted from the publications indexed by PubMed. More than 4000 papers about the Raman spectroscopy in biology and biomedicine were published each year for the past few years.
A most important aspect of the application of Raman technique in medicine is the diagnosis of cancer [4–9]. It is known that most cancers have a good treatment prognosis if an accurate diagnosis is made at an early stage. However, invasive detection methods - biopsy and histology detection - are often unacceptable by patients with early stage cancer because obtaining a biopsy can be painful, the procedure is not without risk of complications, and obvious clinical symptoms are not apparent. In contrast a Raman measurement is a fast, minimally -invasive, painless, repeatable at infinitum, and gives an instantaneous diagnosis, which is of great benefit in the clinic.
Most studies investigating the utility of Raman spectroscopy for clinical applications focused on one organ, where the authors generally compare benign and malignant lesions. For example, the Raman spectroscopic studies of skin cancer demonstrated spectral differences between several malignant and benign lesions, such as basal  and squamous cell carcinoma, lentigo maligna, seborroic keratosis and intradermal nevi . The oral cavity is another common site investigated with a number of reports showing precancer and cancer of the oral mucosa had different Raman features compared to normal mucosa . In vivo Raman spectra also helped to improve the sensitivity and specificity in distinction of precancer tissue and benign tissue of the cervex [12,13]. Raman spectroscopy was also employed to diagnose benign and malignant lesions in human breast tissue ; to identify dysplasia from normal gastric mucosa tissue ; to identify benign, malignant and each pathological group of prostate biopsies ; to differentiate accurately normal and tumoural renal tissue, low grade and high-grade renal tumours ; and to show the difference between normal and malignant lung tissues .
Conventional disease study often starts from an animal model, so it is for the technique of Raman spectroscopy for biology and biomedicine. For example, pancreatic cancer in a mouse model was evaluated with Raman spectroscopy and the difference between cancer and normal tissue was investigated . Different tissue components of the eye in an animal model were also distinguished with Raman microscopy . Furthermore breast tumor tissue could be distinguished from normal breast tissue in a mouse model by Raman spectroscopy, and early neoplastic changes were detected prior to definite morphologic alteration . Those studies on animals were helpful to understand the Raman spectral feature of diseases, and improve the process of clinical application. To date there has been no systematic study of the Raman spectra of different organs.
The features of Raman spectra of different organs should be helpful in the choice of equipment and spectral processing strategies. As a minimally-invasive optical technique, Raman spectroscopy has a good future in becoming a clinical diagnostic guidance technique. The in vivo application demands compact, flexible equipment and fast measurement. Obviously, obtaining a high quality spectrum is hard in such conditions and thus it is very important to match equipment to the spectral characteristics of the intended target.
There are challenges need to be addressed in order to improve the accuracy of Raman spectroscopy, one of them is the interference of sample autofluorescence. For example in our experience, the Raman signal of splanchnic organs is overwhelmed by autofluorescence, making the differences between malignant and benign lesions are more difficult to discern. Although most tissue Raman spectroscopy work focused in the low frequency, fingerprint range, some researchers recorded the Raman signal in high frequency between 1500 and 3500 cm−1 in order to avoid the autofluorescence. And several studies have demonstrated that the high frequency spectrum also contains biochemical information for identification and characterization of tissues [22,23]. In this investigation, we configured our home made Raman system to measure Raman spectra in both low frequency and frequency ranges, covering 700 – 3100 cm-1 in order to completely characterizing the Raman emission of different organs in mice. We hope that the results will be useful in guiding Raman system design and in tissue Raman study.
2. Method and material
2.1. Raman equipment
A schematic of the Raman system we used is shown in Fig. 1 . What follows is a brief description of the system, further details can be found in report of Short et al. . The Raman excitation source was a 785 nm, 500 mW, diode laser (model: BRM-785, B & W Tek, Newark, DE, USA). The mixed tissue fluorescence and Raman signals were analyzed with an f /2 spectrograph equipped with a movable reflection type grating allowing spectra to be obtained from to 0 to 3400 cm−1 (model: LS-785, Princeton Instruments, Trenton, NJ, USA). The intensity of the dispersed light was measured with a thermo-electrically cooled (−70°C) CCD array with 1340x400 pixels (model: PIXIS: 400BR, also from Princeton Instruments). A specially designed catheter consisting of ultra low hydroxide (OH) fibers was used for delivering the excitation light to the tissue surface and for collecting the photons that emerged from the tissue. This catheter was approximately 1.8 mm in diameter and 0.75 m long and consisted of one 200 μm diameter excitation fiber surrounded by 27, 100 μm diameter collection fibers. At the proximal end of the catheter the fibers were separated from the bundle to form excitation and light collection channels that were coupled to a filter module containing collimating lenses and filters. A single 100 μm fiber was used to deliver the excitation light from the laser to the filter module, and a second specially designed fiber bundle guided the filtered emission to the spectrometer. The bundle consisted of 120, ultra low OH, 50 μm diameter fibers packed in a round geometry at the filter module end, but spread out into a parabolic arc and coupled directly to the spectrometer at the other end. This was done to reduce CCD image aberrations which result in an inferior signal to noise ratio and a lower spectral resolution.
Filters placed in the filter module consisted of a laser-line filter (785 ± 2.5 nm) and a long pass interference filter (pass band 800-1200 nm). The laser line filter block spectral contributions from the silica in the excitation fiber and off resonance laser emission. The long pass interference filter attenuated elastically scattered laser light, while allowing passage of fluorescence and Raman scattered light. These filters were obtained from SemRock, Rochester, NY, USA; part numbers LL01-785 and BL01-785 respectively. Another stage of optical filtering was incorporated at the distal end of the catheter by evaporating interference filter coatings directly onto the fibers. For the excitation fiber the coating performed as a short pass filter transmitting λ < 800 nm, and the coating on the collection fibers performed as a long pass filter transmitting λ > 835 nm. Together the two stages of optical filtering substantially reduced elastically scattered light and emissions generated in the fibers from reaching the spectrometer. Control of the system was implemented by a PC using a custom designed program that triggered data acquisition and removed the autofluorescence background in real-time. A low power focusing mode was also incorporated to run the laser at 10% of its maximum power enabling the operator to see the laser spot and thus point the catheter to the location of interest whilst not unduly exposing the tissue to high laser powers. The system was wavelength calibrated, and the estimated spectral resolution was ≈9 cm−1. The spectral intensities were corrected for the system response using a standard halogen calibration lamp.
2.2. Animal and experiment procedure
A popular murine model C3H/HeN was used in this experiment. The experiment was approved by the Animal Care Committee of the University of British Columbia. All mice were housed individually under specific pathogen-free conditions, with a 12 hr light-dark cycle at the animal facility. 6-8 weeks old C3H/HeN mice were supplied by Simonsen Laboratories (Gilroy, CA) and fed two weeks in our Animal Research Center before experiment started. The mice were sacrificed by CO2 following the Standard Operation Procedure of the Animal Research Center. The blood was taken by heart centesis, and the parts of the blood were separated into serum and blood pellets by centrifuging 20 min after taken. A piece of tissue about 5 × 5 × 3 mm3 was cut from each organ and the excess blood absorbed by tissue paper. The organ tissue was then put on a glass slide covered with a standard black electrical tape, which had no measurable Raman emission of its own. No additional processing of the organ tissue was performed and measurements were completed within 10 min. The low frequency spectrum was measured first, by adjusting the distance from the fiber to the tissue surface until the highest intensity spectrum was obtained, then we fixed the fiber position and recorded two spectra. The spectrometer was then adjusted to the high frequency range, and two more spectra were obtained.
2.3. Data processing
The raw data from low frequency (700-2000 cm−1) and high frequency ranges (1900-3100 cm−1) overlapped between 1900 and 2000 cm−1, and this overlap was used to join spectra obtained over the two ranges together into one continuous spectrum. Software named the “Vancouver Raman Algorithm” was then used on the joined spectra to remove the fluorescence background and get Raman spectra . The spectra were subsequently normalized by the area under the curve, and then the mean spectra from individual organs and blood of 5 mice were calculated. Peak positions were identified in the Raman spectra using the “Peak Find” tool in Origin graphing software (OriginLab Corporation, Northampton, MA 01060)
Spectra of eighteen different organs and blood showed different Raman peaks and intensities (Table 1 ). These spectra were divided into 6 groups to compare their features according to their similarity on spectral shape.
3.1. Raman spectra of blood (Fig. 2)
Blood is a specialized body fluid that delivers necessary substances to the cells in the body– such as nutrients and oxygen – and transports waste products away from those same cells. It is composed of blood cells suspended in a liquid called blood plasma. Plasma, which constitutes 55% of blood fluid, is mostly water (90% by volume) and contains dissolved proteins, glucose, mineral ions, hormones, and carbon dioxide. The blood cells are mainly red blood cells (erythrocytes) and white blood cells, including leukocytes and platelets. The red blood cells contain hemoglobin, an iron-containing protein.
The spectrum of whole blood was almost the same as the blood pellet. This suggests that the spectrum is mainly come from red blood cells. The major Raman peaks were located around 742, 778, 991, 1074, 1120, 1160, 1210, 1335, 1383, 1442, 1542, 1614, 2159, 2914 cm−1. Deng and coauthors reported the Raman spectrum of single blood cells in the low frequency range with similar results to ours . Another report about the measurement of hemoglobin oxygen saturation showed that peaks of 1375, 1590, 1640 cm−1 would increase with high O2 saturation , these peaks did not appear in this experiment as the O2 saturation of blood was low. The small sharp peak located near 2330 cm−1 is probably that of atmospheric N2 which occurs because of a high excitation power density and large collection solid angle near the surface of fibers at the distal end of the catheter.
Serum is the supernatant fluid when coagulated blood has been centrifuged, it is devoid of all coagulation factors comparing with plasma. The spectrum of serum without blood cells was quite different from that of whole blood in the low frequency range. The major Raman peaks were located around 820, 1044, 1335, 1383, 1442, 1542, 1614, 1653, 2159, 2646, 2914cm−1.
All the three spectra showed similar peaks at 1335, 1383, 1442, and 1614 cm−1 which mainly come from CH2, CH3 (lipids and proteins). In high frequency, the dominant peak is at 2914 cm−1 which mainly come from the protein component. The 1044 and 1653 cm−1 peaks appeared in the spectrum of the serum, which also showed up in the spectra of the stomach, small intestine, colon, bladder, lung, and brain in data to be presented later in the paper, but did not appear in the spectra of whole blood and blood pellet.
3.2. Raman spectra of adipose tissue, skeletal muscle, and skin (Fig. 3)
Adipose tissue is loose connective tissue composed of adipocytes. Its main role is to store energy in the form of fat. Adipose tissue also serves as an important endocrine organ by producing hormones such as leptin, resistin, and the cytokine TNFα. The samples, which come from fat tissue inside the abdomen, contains several cell types, with the highest percentage of cells being adipocytes, which contain fat droplets. Other cell types include fibroblasts, macrophages, and endothelial cells. Biochemically adipose tissue is composed of lipids, cholesterol, fatty acid, unsaturated fatty acid, and cholesterol ester. The characteristic peaks of adipose tissue were located in 1065, 1270, 1298, 1437, 1650 cm−1 in the low frequency region, and 2828, 2879, 2970 cm−1 in the high frequency region. The Raman shift peaks in the low frequency region is consistent with report by other studies , and to the best of our knowledge there is no report of adipose tissue Raman spectra in the high frequency region. The adipose tissue spectra have strong similarities to a palmitic acid spectrum . Palmitic acid is one of the most common saturated fatty acids found in animals and plants, it mainly occurs as its ester in triglycerides (fats). In fact, all saturated fatty acids including lauric acid, myristic acid, palmitic acid and stearic acid have similar Raman spectra with strong peak in 1063, 1128, 1296, 1438 cm−1 .
Skeletal muscles are the contractile tissue. Muscle cells contain contractile filaments that move past each other and change the size of the cell. Their function is to produce force and cause motion. Voluntary contraction of the skeletal muscles is used to move the body and can be finely controlled. Muscle is mainly composed of muscle cells. Within the cells are myofibrils containing sarcomeres, which are composed of actin and myosin. Muscular activity accounts for much of the body's energy consumption. All muscle cells produce adenosine triphosphate (ATP) molecules which are used to power the movement of the myosin heads. Muscles also keep a storage form of glucose in the form of glycogen and contain globules of fat, which provides energy during aerobic exercise. The characteristic peaks of skeletal muscles are located in 851, 962, 1065, 1258, 1297, 1437, 1542, 1653 and 1737cm−1 in the low frequency region, and 2159, 2698, 2828, 2914 and 2987 cm−1 in the high frequency region. Up to now, Raman spectroscopy has not yet used in the study of human muscles.
Skin is a very important organ providing functions of protection (an anatomical barrier from pathogens and damage), sensation, heat regulation, control of evaporation and water resistance. Skin is composed of epidermis, dermis, and hypodermis. Epidermis is made up of stratified squamous epithelium with an underlying basal lamina. Dermis is the layer of skin beneath the epidermis that consists of connective tissues. It harbors many nerve endings that provide the sense of touch and heat, and contains hair follicles, sweat glands, sebaceous glands, apocrine glands, lymphatic vessels and blood vessels. The hypodermis is not part of the skin and lies below the dermis. Its purpose is to attach the skin to underlying bone and muscle. It consists of loose connective tissue and elastin. The main cell types are fibroblasts, macrophages and adipocytes (the hypodermis contains 50% of body fat). The major Raman peaks of skin are 851, 962, 1065, 1258, 1297, 1437, 1542, 1653, 1737, 2159, 2698, 2828, 2879 and 2987 cm−1, similar to reference .
The Raman spectrum of ventral muscle and skin were very similar to the adipose tissue. The characteristic peaks at 1065, 1258, 1297, 1437, 1653, 2828, 2987 cm−1 of adipose tissue also appeared in the Raman spectra of ventral muscle and skin. One reason is that some peaks could come from both fat and protein, just like the one at 1653 cm−1, and the 1258 cm−1 peak could come from protein, lipid and guanine, cytosine. Another reason maybe that the fat in its solitary state exists in the muscle cells, and a high concentration of lipids is found in whole skin, especially in the hypodermis . As fat has stronger Raman signal, it contributes significantly to muscle and skin Raman spectra. More work is needed to fully understand the similarity in ventral muscle and fat spectra.
Other peaks like the one at 1542 cm−1 which comes from proteins only appeared in muscle.
3.3. Raman spectra of the gastrointestinal tract (Fig. 4)
Gastrointestinal tract including stomach, small intestine and colon have similar structures. The stomach is a muscular, hollow, dilated digestive tract. It secretes protein-digesting enzymes and strong acids to aid in food digestion. Like the other parts of the gastrointestinal tract, the stomach walls are made of the following layers, from surface to underneath: The first main layer is mucosa which consists of an epithelium, gastric glands, loose connective tissue and the muscularis mucosae. The second layer is submucosa which consists of fibrous connective tissue. Below the submucosa there is muscularis externa which has three layers of smooth muscles. The forth layer is serosa consisting of layers of connective tissue continuous with the peritoneum. The characteristic Raman peaks of normal stomach tissue are 828, 851, 991, 1044, 1258, 1302, 1442, 1653, 1725, 2139, 2177 and 2917 cm−1. Similar peaks in the low frequency region have also been reported by other groups , and there is no report about Raman spectrum of stomach in the high frequency region.
The small intestine is the part of the gastrointestinal tract following the stomach and followed by the large intestine (or colon), and is where the vast majority of digestion and absorption of food takes place. It also consists of mucosa, submucosa, muscularis externa, and serosa. The mucosa of small intestine has intestinal epithelium consisting of simple columnar with very long villi and the microvilli, serving to increase the amount of surface area available for the absorption of nutrients. The characteristic Raman peaks of normal small intestine tissue are 828, 921, 991, 1044, 1074, 1160, 1258, 1302, 1335, 1442, 1542, 1653, 1725, 2139, 2177, 2870 and 2917 cm−1. There is still no report about the Raman spectrum of small intestine.
The colon is the last part of the digestive system; it is mainly responsible for storing waste, reclaiming water, maintaining the water balance, absorbing vitamins, such as vitamin K. The Raman spectrum of colon is similar to that of small intestine, even the position and intensity of the peaks are the same. Andrade et al  reported Raman spectra of normal colorectal tissue with peaks at 1080, 1260, 1300, 1450, 1650 and 1750 cm−1, which also appeared in our colon spectrum shown in Fig. 4 ; their study did not include measurements in the high frequency region.
These three organs have the same peaks at 828, 1258, 1302, 1442, 1653, 1725, 2136, 2169 and 2917 cm−1. The peaks at 828, 1258 cm−1 come from protein and DNA; whereas the peaks at 1302, 1442, 1653, and 2917 cm−1 come from lipid and protein. The peak at 1725 cm−1 come from lipid, and those at 2136, 2169 cm−1 come from water, which appeared in many organs. Small intestine and colon have more peaks at 921, 1074, 1160, 1335, 1542, 2870 cm−1 than stomach.
3.4. Raman spectra of urinary bladder, lung, and brain (Fig. 5)
The urinary bladder is the organ that collects urine excreted by the kidneys prior to disposal by urination. It is also a hollow muscular and elastic organ, and composed of mucosa, submucosa, smooth muscle and serosa. The mucosa of bladder consists of transitional epithelium without glands. The characteristic Raman peaks of bladder are 828, 991, 1044, 1258, 1302, 1442, 1542, 1614, 1653, 1725, 2139 and 2917 cm−1. The urinary bladder’s Raman spectrum was similar to the gastrointestinal tract. They all have same site and intensity at 828, 991, 1044, 1442, 1653, 1725, 2139 and 2917 cm−1. The intensity of 1258, 1302, 1542 cm−1 in bladder is weaker than that in gastrointestinal tract. The reason maybe that bladder have similar structure of wall to gastrointestinal tract.
The lung is the essential respiration organ. Their principal function is to transport oxygen from the atmosphere into the bloodstream, and to release carbon dioxide from the bloodstream into the atmosphere. The bronchial tree continues branching until it reaches the level of terminal bronchioles, which lead to alveolar sacs. The individual alveoli are tightly wrapped in blood vessels and it is here that gas exchange actually occurs. The lungs have a spongy texture. So the lung sample included bronchioles, alveoli and blood vessels. The characteristic Raman peaks of lung are 800, 991, 1044, 1302, 1335, 1442, 1542, 1590, 1614, 1653, 1725, 2139 and 2917 cm−1. Lung spectra had a special peak at 1590 cm−1, It is unclear what the cause of this peak is; carbon particles in human specimen  where found to give a strong 1590 cm−1 peak, although they were accompanied by a 1350 cm−1 peak which was not observed in our spectra. Another possibility is that euthanizing mice with CO2 results in some unknown biochemical changes in the lungs that result in a strong 1590 cm−1 peak. Strong peaks of 1302, 1445, 1655 cm−1 in the low frequency region were similar to other report, but the weak peaks were different . The Raman spectrum of normal and malignant human lung tissue in the high frequency region has been investigated by Short et al .
The brain is the center of the nervous system. It is composed of two broad classes of cells: neurons and glia. The cerebral cortex of the human brain contains roughly 15–33 billion neurons. These neurons communicate with one another by axons, which carry trains of signal pulses called action potentials to distant parts of the brain or body and target them to specific recipient cells. Glia comes in several types, which perform a number of critical functions, including structural support, metabolic support, insulation, and guidance of development. The primary function of a brain is to control the actions of an animal. The brain sample was the cerebral cortex which is a layer of gray matter that lies on the surface of the forebrain. The characteristic Raman peaks of brain are 962, 991, 1044, 1302, 1442, 1542, 1614, 1653, 1725, 2139, 2879 and 2917 cm−1. Brain tissue has a strong peak at 2879 cm−1, which also appeared in skin and adipose tissue and comes from lipids and proteins. Considering lipids are so abundant in brain tissue, this peak is most likely originated from lipids. Lakshmi et al  compared the Raman spectra between brain gray matter to white matter, the main peaks in the 1000-1700 cm−1 range were similar to our results. In the high frequency region, peaks at 2879 and 2917 cm−1 reported by this group were appeared in our results, but not the ones at 2856, 2899, 2926 cm−1. The reason is not clear.
3.5. Raman spectra of kidney, heart, liver and spleen (Fig. 6)
Kidneys serve the body as natural filters of the blood, and for the removal of waste. They also serve homeostatic functions such as the regulation of electrolytes, maintenance of acid-base balance, and regulation of blood pressure. The kidney is divided into two major structures: superficial is the renal cortex and deep is the renal medulla. The sample was taken from renal cortex. The peaks in the Raman spectra of kidney tissue were: 876, 1031, 1335, 1442, 1542, 1623, 1653, 1725, 2127, 2159 and 2914 cm−1. Bensalah et al  reported different Raman spectra of normal and tumoural renal tissue, low grade and high-grade renal tumours, and histologic subtype of renal cell carcinoma. But they didn’t measure the Raman spectra in the high frequency range.
The heart is a muscular organ that is responsible for pumping blood throughout the blood vessels by repeated, rhythmic contractions. It is mainly composed of cardiac muscle, which is an involuntary striated muscle tissue found only within heart. The outer wall of heart is composed of three layers. The outer layer is called the epicardium. The middle layer is called the myocardium and is composed of muscle which contracts. The inner layer is called the endocardium and is in contact with the blood that the heart pumps. The peaks in the Raman spectra of heart tissue were: 962, 1031, 1302, 1335, 1442, 1542, 1623, 1653, 1725, 2127, 2159, 2870 and 2914 cm−1. There is no report about the Raman spectra of heart yet.
Liver is necessary for survival with wide range of functions, including detoxification, protein synthesis, and production of biochemicals necessary for digestion. The liver's highly specialized tissues regulate a wide variety of high-volume biochemical reactions, including the synthesis and breakdown of small and complex molecules, many of which are necessary for normal vital functions. The bile produced in the liver is collected in bile canaliculi, which merge to form bile ducts. Liver is made up of lobules composed of liver cells or hepatocytes. The peaks in the Raman spectra of liver tissue were: 851, 1044, 1335, 1442, 1542, 1585, 1623, 1653, 1725, 2159, 2870 and 2914 cm−1. Shen et al  investigated using Raman spectra to distinguish alcoholic liver injury and ethanol-induced liver fibrosis from the normal state. The peaks of normal liver in the low frequency region were similar to our results, but the peaks in high frequency region were not investigated.
Spleen has important roles in regard to red blood cells and the immune system. It is one of the centers of activity of the reticuloendothelial system. Spleen is composed of red pulp performing filtration of red blood cells and reserve of monocytes, and white pulp, which could have immune response through humoral and cell-mediated pathways. The peaks in the Raman spectra of spleen tissue were: 828, 1017, 1442, 1542, 1623, 1725, 2127, 2151, 2747 and 2914 cm−1. No investigation of Raman spectra of spleen have been reported.
3.6 Raman spectra of teeth and skull (Fig. 7)
Teeth are used to break down food. It includes enamel, dentin, cementum, and dental pulp. Enamel is the hardest and most highly mineralized substance of the body. It consists of ninety-six percent of hydroxyapatite, which is a crystalline calcium phosphate. Dentin is made up of 70% hydroxyapatite, 20% organic materials, and 10% water by weight. Cementum is approximately 45% inorganic material (mainly hydroxyapatite), 33% organic material (mainly collagen) and 22% water. The dental pulp is the central part of the tooth filled with soft connective tissue which contains blood vessels and nerves. The peaks in the Raman spectra of teeth were: 800, 851, 950, 1065, 1335, 1442, 1542, 1653, 1725, 2139 and 2917 cm−1.
Bone function to move, support, and protect the various organs of the body, produce red and white blood cells and store minerals. One type of tissue that makes up bone is the mineralized osseous tissue, also called bone tissue. Other types of tissue found in bones include marrow, endosteum and periosteum, nerves, and blood vessels. Osseous tissue formed mostly of calcium phosphate in the chemical arrangement termed calcium hydroxylapatite. The sample was skull belonging to flat bones which are thin and generally curved, with two parallel layers of compact bones sandwiching a layer of spongy bone. The peaks in the Raman spectra of skull were similar to teeth, and the sites were: 800, 851, 950, 1065, 1335, 1442, 1542, 1653, 1725, 2139 and 2917 cm−1.
Both tooth and skull have a strong Raman peak at 950 cm−1 coming from calcium hydroxylapatite , the intensity of this peak in skull was weaker than that in tooth as there were more calcium hydroxylapatite in teeth. Other peaks like 851, 1442 cm−1 come from proteins, and the 1065, 1653 cm-1 come from lipids. The peaks at 1442, 1653 cm−1 had been used to investigate the mechanical deformation of bone tissues [34,35].
4.1. The influence of tissue compositions on Raman spectrum
We have shown clear differences in the spectra from different organs and tissue, and these differences will be present in spectra acquired in vivo as well, although there may be a larger blood contribution for some samples. In general, different molecules or different chemical bond structure should have a unique characteristic Raman spectrum. Although the main peaks coming from protein, lipid and DNA appear in similar positions in Raman spectra of all tissues because all tissues have protein molecules, phospholipids molecules, DNA and RNA. For example, the peaks at 1270, 1310, 1445, 1660, and 2900 cm−1 originated from lipids and proteins are very clear in every organ.
However, if an organ has big proportion of a unique substance, characteristic peaks of this molecule will appear in the spectrum of this organ. For example, the brain has a strong peak at 2879 cm−1, which comes from lipid. The most likely reason is that brain contains a large amount of lipid. In contrast other biological tissue is unique in that it has certain Raman peaks missing, for example red blood cells do not have peaks at 785 and1090 cm−1 originated from DNA  because of the lack of nuclear materials in erythrocytes.
In addition some tissues have special elements, for example teeth and bone have calcium and therefore, a strong peak at 950 cm−1 which had been employed as an analytical tool in the detection of basic calcium phosphate in the joint fluid of osteoarthritis patients .
Other tissues, such as fat have strong and clear Raman signals. If an organ has fat tissue, Raman signals from other compositions can be overwhelmed by the intense Raman signals of fatty tissue. This makes it difficult to determine changes that may occur in other compositions by monitoring changes in their characteristic Raman peaks. In this situation, a defined micro-volume detected with confocal Raman spectroscopy is a good way to study [38,39]. Furthermore it should be noted that the organ spectra presented here are average spectra representing the average for the whole organ, spatial variations within an organ are to be expected. Some within organ variance in the spectra was observed, but a more in-depth study would be required to determine if this was real difference between the same organ from different mice or simply a spatial variance.
4.2. Fluorescence background issue
We found that autofluorescence background influence the measurement of Raman spectra of organs. The main fluorophores in biological tissue are pyridinic (NADPH) and flavin coenzymes (FAD), collagen and elastin. In Raman measurements, fluorescence signal is always presented, sometimes making it difficult to separate Raman signal from fluorescence background. Because Raman signal is always orders lower than fluorescence, the total measurement time is limited by the intensity of fluorescence. And as the measured Raman signal proportionally depends on the measurement time, once the fluorescence saturates the detector, there is no way to further increase the Raman signal by simply increasing the measurement time. Therefore, knowing the ratio of Raman to fluorescence signal is important in Raman spectroscopy instrumentation design and Raman spectral measurements. One reason that we choose NIR excitation for Raman measurements is that the tissue fluorescence are relatively lower than excited by a shorter wavelength. Based on the relative levels of fluorescence and Raman, we found that the spectra of mouse organs can be divided into three categories: (i) low fluorescence, high Raman signal, (ii) moderate fluorescence, moderate Raman signal, and (iii) high fluorescence, low Raman signal. The third category will present technical challenges in Raman spectral measurements. For example, Fig. 8 shows the raw spectra (mixed fluorescence and Raman) of adipose tissue, colon and liver. It can be seen that adipose tissue belongs to category one with low fluorescence and high Raman signal; colon belongs to category two with moderate fluorescence and moderate Raman signal; and liver belongs to category three with high fluorescence and low Raman signals. The Raman spectra of these three organs after background fluorescence removal using the Vancouver Raman Algorithm  were shown in Fig. 9 . We can see that adipose tissue has strong Raman signal, colon has medium Raman signal and liver has weak Raman signal.
Organs belonging to the first category have very strong Raman signal and weak fluorescence signal in both the high frequency and low frequency region. The Raman spectra of these organs can be easily acquired, and Raman peaks can be easily identified, especially in the low frequency fingerprint region. These organs include: skin, muscle, adipose tissue, whole blood, blood pellet, blood serum, tooth and bone, (See Figs. 2 , 3 and 7 ). They all have low fluorescence and strong Raman spectra. Tooth and bone have a characteristic Raman peak at 950 cm−1 which is assigned to calcium (Fig. 7). Thus low frequency region Raman spectrum might be necessary in studying calcium related tissues.
Organs in category two have moderate Raman signals in both the low-frequency and high-frequency regions. For these organs, either the high frequency or the low frequency region Raman spectra can be measured to quantify the properties of the tissues. These organs include all digestive organs such as stomach, small intestine, and colon (Fig. 4), and other organs such as brain, lung, and bladder (Fig. 5 ). These organs have relatively light color.
Organs in the third category have strong fluorescence and extremely low Raman signals in the low frequency range but measurable Raman signals in the high frequency range. For these organs, it is difficult to measure the Raman spectra in the low frequency region. Only the high frequency Raman spectra is usable. These organs include spleen, kidney, liver and heart (Fig. 6 ), usually appear dark red. To get acceptable Raman spectra of these organs, other enabling technique has to be used, such as applying Au or Ag nanoparticles to facilitate surface-enhanced Raman spectroscopy measurements .
Raman spectroscopy is a helpful technique in diseases diagnosis. The characteristic of Raman spectra is a key point for experiment design and investigation. In this paper, for the first time Raman spectra from 18 normal murine organs were measured in both the low frequency and the high frequency regions. And Raman properties of these organs are quantified and characterized. We found that spectra are similar if the structures of organs are similar, for example, the Raman spectra of the stomach, small intestine, colon and bladder are similar because they are all composed of three similar layers and have large amount of muscle. Both bone and teeth have great amount of calcium, and thus have very similar Raman spectra. Most Raman spectra have peaks at 991, 1044, 1258, 1335, 1442, 1542, 1653, 1725, 2914 cm−1, which were assigned to proteins and lipids, but the relative intensities were different in each curve as the relative intensities have relationships with the proportion of proteins and lipids in the organ.
Each Raman spectrum has unique features, for example, peaks at 1017, 2151, 2747 cm−1 only appeared in the spectrum of spleen; peak at 876 cm−1 appeared in the spectrum of kidney, peak at 921 cm−1 appeared in the spectra of both small intestine and colon; and 1031 cm−1 appeared in the spectra of both kidney and heart.
Raman properties of these organs were compared with their fluorescence counterparts. We found that the spectra of mouse organs can be divided into three categories: (i) low fluorescence and high Raman signal, (ii) moderate fluorescence and moderate Raman signal, and (iii) high fluorescence and low Raman signal. This provides direction in designing Raman probes and Raman systems for in vivo applications.
This work was financially supported by the Canadian Dermatology Foundation. We are grateful to Dr. David I McLean for helpful discussions during the project. And we thanks Mr. Wei Zhang, Mr. Soroush Merchant for their technical assistance.
Reference and links:
3. E. B. Hanlon, R. Manoharan, T.-W. Koo, K. E. Shafer, J. T. Motz, M. Fitzmaurice, J. R. Kramer, I. Itzkan, R. R. Dasari, and M. S. Feld, “Prospects for in vivo Raman spectroscopy,” Phys. Med. Biol. 45(2), R1–R59 (2000). [CrossRef] [PubMed]
4. O. R. Sćepanović, Z. Volynskaya, C.-R. Kong, L. H. Galindo, R. R. Dasari, and M. S. Feld, “A multimodal spectroscopy system for real-time disease diagnosis,” Rev. Sci. Instrum. 80(4), 043103 (2009). [CrossRef] [PubMed]
5. M. Y. Sha, H. Xu, M. J. Natan, and R. Cromer, “Surface-enhanced Raman scattering tags for rapid and homogeneous detection of circulating tumor cells in the presence of human whole blood,” J. Am. Chem. Soc. 130(51), 17214–17215 (2008). [CrossRef] [PubMed]
6. C. A. Lieber, S. K. Majumder, D. L. Ellis, D. D. Billheimer, and A. Mahadevan-Jansen, “In vivo nonmelanoma skin cancer diagnosis using Raman microspectroscopy,” Lasers Surg. Med. 40(7), 461–467 (2008). [CrossRef] [PubMed]
7. Z. Huang, H. Lui, D. I. McLean, M. Korbelik, and H. Zeng, “Raman spectroscopy in combination with background near-infrared autofluorescence enhances the in vivo assessment of malignant tissues,” Photochem. Photobiol. 81(5), 1219–1226 (2005). [CrossRef] [PubMed]
9. G. Shetty, C. Kendall, N. Shepherd, N. Stone, and H. Barr, “Raman spectroscopy: elucidation of biochemical changes in carcinogenesis of oesophagus,” Br. J. Cancer 94(10), 1460–1464 (2006). [CrossRef] [PubMed]
10. A. Nijssen, T. C. Bakker Schut, F. Heule, P. J. Caspers, D. P. Hayes, M. H. A. Neumann, and G. J. Puppels, “Discriminating basal cell carcinoma from its surrounding tissue by Raman spectroscopy,” J. Invest. Dermatol. 119(1), 64–69 (2002). [CrossRef] [PubMed]
11. R. Malini, K. Venkatakrishna, J. Kurien, K. M. Pai, L. Rao, V. B. Kartha, and C. M. Krishna, “Discrimination of normal, inflammatory, premalignant, and malignant oral tissue: a Raman spectroscopy study,” Biopolymers 81(3), 179–193 (2006). [CrossRef] [PubMed]
12. M. D. Keller, E. M. Kanter, C. A. Lieber, S. K. Majumder, J. Hutchings, D. L. Ellis, R. B. Beaven, N. Stone, and A. Mahadevan-Jansen, “Detecting temporal and spatial effects of epithelial cancers with Raman spectroscopy,” Dis. Markers 25(6), 323–337 (2008). [PubMed]
13. C. Murali Krishna, G. D. Sockalingum, M. S. Vidyasagar, M. Manfait, D. J. Fernanades, B. M. Vadhiraja, and K. Maheedhar, “An overview on applications of optical spectroscopy in cervical cancers,” J. Cancer Res. Ther. 4(1), 26–36 (2008). [CrossRef] [PubMed]
14. A. S. Haka, K. E. Shafer-Peltier, M. Fitzmaurice, J. Crowe, R. R. Dasari, and M. S. Feld, “Diagnosing breast cancer by using Raman spectroscopy,” Proc. Natl. Acad. Sci. U.S.A. 102(35), 12371–12376 (2005). [CrossRef] [PubMed]
15. S. K. Teh, W. Zheng, K. Y. Ho, M. Teh, K. G. Yeoh, and Z. Huang, “Diagnostic potential of near-infrared Raman spectroscopy in the stomach: differentiating dysplasia from normal tissue,” Br. J. Cancer 98(2), 457–465 (2008). [CrossRef] [PubMed]
16. P. Crow, N. Stone, C. A. Kendall, J. S. Uff, J. A. Farmer, H. Barr, and M. P. Wright, “The use of Raman spectroscopy to identify and grade prostatic adenocarcinoma in vitro,” Br. J. Cancer 89(1), 106–108 (2003). [CrossRef] [PubMed]
17. K. Bensalah, J. Fleureau, D. Rolland, O. Lavastre, N. Rioux-Leclercq, F. Guillé, J.-J. Patard, L. Senhadji, and R. de Crevoisier, “Raman spectroscopy: a novel experimental approach to evaluating renal tumours,” Eur. Urol. 58(4), 602–608 (2010). [CrossRef] [PubMed]
18. Z. Huang, A. McWilliams, H. Lui, D. I. McLean, S. Lam, and H. Zeng, “Near-infrared Raman spectroscopy for optical diagnosis of lung cancer,” Int. J. Cancer 107(6), 1047–1052 (2003). [CrossRef] [PubMed]
19. A. K. Pandya, G. K. Serhatkulu, A. Cao, R. E. Kast, H. Dai, R. Rabah, J. Poulik, S. Banerjee, R. Naik, V. Adsay, G. W. Auner, M. D. Klein, J. S. Thakur, and F. H. Sarkar, “Evaluation of pancreatic cancer with Raman spectroscopy in a mouse model,” Pancreas 36(2), e1–e8 (2008). [CrossRef] [PubMed]
20. N. Terada, N. Ohno, S. Saitoh, Y. Fujii, H. Ohguro, and S. Ohno, “Raman microscopy of freeze-dried mouse eyeball-slice in conjunction with the “in vivo cryotechnique”,” Microsc. Res. Tech. 70(7), 634–639 (2007). [CrossRef] [PubMed]
21. R. E. Kast, G. K. Serhatkulu, A. Cao, A. K. Pandya, H. Dai, J. S. Thakur, V. M. Naik, R. Naik, M. D. Klein, G. W. Auner, and R. Rabah, “Raman spectroscopy can differentiate malignant tumors from normal breast tissue and detect early neoplastic changes in a mouse model,” Biopolymers 89(3), 235–241 (2008). [CrossRef] [PubMed]
22. K. W. Short, S. Carpenter, J. P. Freyer, and J. R. Mourant, “Raman spectroscopy detects biochemical changes due to proliferation in mammalian cell cultures,” Biophys. J. 88(6), 4274–4288 (2005). [CrossRef] [PubMed]
24. M. A. Short, S. Lam, A. McWilliams, J. Zhao, H. Lui, and H. Zeng, “Development and preliminary results of an endoscopic Raman probe for potential in vivo diagnosis of lung cancers,” Opt. Lett. 33(7), 711–713 (2008). [CrossRef] [PubMed]
25. J. Zhao, H. Lui, D. I. McLean, and H. Zeng, “Automated autofluorescence background subtraction algorithm for biomedical Raman spectroscopy,” Appl. Spectrosc. 61(11), 1225–1232 (2007). [CrossRef] [PubMed]
26. J. L. Deng, Q. Wei, M. H. Zhang, Y. Z. Wang, and Y. Q. Li, “Study of the effect of alcohol on single human red blood cells using near-infrared laser tweezers Raman spectroscopy,” J. Raman Spectrosc. 36(3), 257–261 (2005). [CrossRef]
27. I. P. Torres Filho, J. Terner, R. N. Pittman, E. Proffitt, and K. R. Ward, “Measurement of hemoglobin oxygen saturation using Raman microspectroscopy and 532-nm excitation,” J. Appl. Physiol. 104(6), 1809–1817 (2008). [CrossRef] [PubMed]
28. M. Gniadecka, O. Faurskov Nielsen, D. H. Christensen, and H. C. Wulf, “Structure of water, proteins, and lipids in intact human skin, hair, and nail,” J. Invest. Dermatol. 110(4), 393–398 (1998). [CrossRef] [PubMed]
29. P. O. Andrade, R. A. Bitar, K. Yassoyama, H. Martinho, A. M. E. Santo, P. M. Bruno, and A. A. Martin, “Study of normal colorectal tissue by FT-Raman spectroscopy,” Anal. Bioanal. Chem. 387(5), 1643–1648 (2007). [CrossRef] [PubMed]
30. Y.-K. Min, T. Yamamoto, E. Kohda, T. Ito, and H. Hamaguchi, “1064 nm near-infrared multichannel Raman spectroscopy of fresh human lung tissues,” J. Raman Spectrosc. 36(1), 73–76 (2005). [CrossRef]
31. R. Jyothi Lakshmi, V. B. Kartha, C. Murali Krishna, J. G. R Solomon, G. Ullas, and P. Uma Devi, “Tissue Raman spectroscopy for the study of radiation damage: brain irradiation of mice,” Radiat. Res. 157(2), 175–182 (2002). [CrossRef] [PubMed]
32. A. Shen, B. Zhang, J. Ping, W. Xie, P. Donfack, S.-J. Baek, X. Zhou, H. Wang, A. Materny, and J. Hu, “In vivo study on the protection of indole-3-carbinol (I3C) against the mouse acute alcoholic liver injury by micro-Raman spectroscopy,” J. Raman Spectrosc. 40(5), 550–555 (2009). [CrossRef]
33. L. Silveira Jr, S. Sathaiah, R. A. Zângaro, M. T. T. Pacheco, M. C. Chavantes, and C. A. G. Pasqualucci, “Correlation between near-infrared Raman spectroscopy and the histopathological analysis of atherosclerosis in human coronary arteries,” Lasers Surg. Med. 30(4), 290–297 (2002). [CrossRef] [PubMed]
34. J. A. Timlin, A. Carden, M. D. Morris, R. M. Rajachar, and D. H. Kohn, “Raman spectroscopic imaging markers for fatigue-related microdamage in bovine bone,” Anal. Chem. 72(10), 2229–2236 (2000). [CrossRef] [PubMed]
35. A. Carden, R. M. Rajachar, M. D. Morris, and D. H. Kohn, “Ultrastructural changes accompanying the mechanical deformation of bone tissue: a Raman imaging study,” Calcif. Tissue Int. 72(2), 166–175 (2003). [CrossRef] [PubMed]
37. A. Yavorskyy, A. Hernandez-Santana, G. McCarthy, and G. McMahon, “Detection of calcium phosphate crystals in the joint fluid of patients with osteoarthritis - analytical approaches and challenges,” Analyst (Lond.) 133(3), 302–318 (2008). [CrossRef] [PubMed]
39. H. Wang, N. Huang, J. Zhao, H. Lui, M. Korbelik, and H. Zeng, “Depth-resolved in vivo micro-Raman spectroscopy of a murine skin tumor model reveals cancer-specific spectral biomarkers,” J. Raman Spectrosc. 42(2), 160–166 (2011). [CrossRef]
41. J. W. Chan, D. S. Taylor, T. Zwerdling, S. M. Lane, K. Ihara, and T. Huser, “Micro-Raman spectroscopy detects individual neoplastic and normal hematopoietic cells,” Biophys. J. 90(2), 648–656 (2006). [CrossRef] [PubMed]
42. N. Stone, C. Kendall, N. Shepherd, P. Crow, and H. Barr, “Near-infrared Raman spectroscopy for the classification of epithelial pre-cancers and cancers,” J. Raman Spectrosc. 33(7), 564–573 (2002). [CrossRef]
46. I. Notingher, C. Green, C. Dyer, E. Perkins, N. Hopkins, C. Lindsay, and L. L. Hench, “Discrimination between ricin and sulphur mustard toxicity in vitro using Raman spectroscopy,” J. R. Soc. Interface 1(1), 79–90 (2004). [CrossRef] [PubMed]
48. Z. Huang, A. McWilliams, S. Lam, J. English, D. I. McLean, H. Lui, and H. Zeng, “Effect of formalin fixation on the near-infrared Raman spectroscopy of normal and cancerous human bronchial tissues,” Int. J. Oncol. 23(3), 649–655 (2003). [PubMed]
49. N. J. Kline and P. J. Treado, “Raman Chemical Imaging of Breast Tissue,” J. Raman Spectrosc. 28(2–3), 119–124 (1997). [CrossRef]
50. J. De Gelder, K. De Gussem, P. Vandenabeele, and L. Moens, “Reference database of Raman spectra of biological molecules,” J. Raman Spectrosc. 38(9), 1133–1147 (2007). [CrossRef]