Biochemical techniques: centrifugation and spectrophotometry

Biochemistry (B.Sc) · Techniques · free study notes for exams, placements and interviews

In short

Centrifugation separates particles in a liquid by spinning them so that a large artificial gravitational field (hundreds to a million times g) drives sedimentation at rates set by size, shape and density. Spectrophotometry measures how much light of a chosen wavelength a solution absorbs, and converts that into concentration through the Beer-Lambert law. Together they are the backbone of every biochemistry laboratory: cells are fractionated, proteins and nucleic acids pelleted or banded, and then quantified or assayed on a UV-Vis spectrophotometer. Both are standard university practical and viva topics, appear in GATE (BT/XL), CSIR-UGC NET life sciences, ICMR/DBT-JRF and ARS-NET, and are the first questions asked in QC and R&D interviews at Indian employers such as Serum Institute, Bharat Biotech, Biocon, Dr. Reddy's, Cipla, Syngene, SRL and Thyrocare.

Explained

Principle of centrifugation

A particle spun at angular velocity ω at radius r experiences a centrifugal field ω²r; expressed as a multiple of gravity it is the relative centrifugal force (RCF or "× g").
Sedimentation of a sphere in a viscous fluid follows Stokes' law: velocity rises with the square of the diameter and with the density difference (ρp − ρm), and falls with medium viscosity η.
If ρp = ρm the particle does not move; if ρp < ρm it floats (the basis of flotation of lipoproteins and of isopycnic banding).
Sedimentation coefficient s = velocity per unit field, in Svedberg units (1 S = 10⁻¹³ s). Typical values: ribosomal subunits 30S/50S (prokaryote) and 40S/60S (eukaryote), intact ribosomes 70S/80S, lysozyme ~1.9 S, catalase ~11.3 S.
s values are not additive: 30S + 50S gives 70S, not 80S, because shape and hydration change on assembly.
Sedimentation is always quoted as g and time (e.g. 10,000 × g for 15 min), never as rpm alone, because the same rpm gives different g in different rotors.

Instrument, rotors and practice

Parts: motor/drive, rotor chamber (often refrigerated and under vacuum in ultracentrifuges), rotor, speed and temperature controls, imbalance and lid-lock interlocks, brake/deceleration profile.
Classes: clinical/benchtop (up to ~5,000 rpm, a few thousand g - blood, urine); microcentrifuge (up to ~15,000 rpm, ~21,000 × g); high-speed refrigerated (~20,000-25,000 rpm, up to ~60,000 × g); preparative ultracentrifuge (up to ~150,000 rpm, ~1,000,000 × g); analytical ultracentrifuge with optical detection; continuous-flow centrifuges and industrial disc-stack separators for harvest at plant scale.
Rotor types:
- Fixed-angle (typically 14-45°): short sedimentation path, fast pelleting, pellet forms on the tube wall - best for differential centrifugation.
- Swinging-bucket: tubes swing to horizontal; longest path, poorer pelleting but the best resolution for density-gradient work (bands stay flat and horizontal).
- Vertical and near-vertical: very short path, fastest gradient runs (CsCl plasmid preps), but gradients must be reoriented.
- Zonal and continuous-flow rotors for large volumes (vaccine and virus production).
k-factor (clearing factor) rates a rotor at its maximum speed: pelleting time t (h) = k/s, so a lower k means faster clearing.
Practice: balance tubes by weight to within the manufacturer's limit; never exceed the rotor's rated speed; derate the rotor for dense media (CsCl) and for ageing/fatigue as per the log book; use caps for infectious or volatile samples; a cracked or corroded tube or rotor is scrapped, not used.
Safety: rotor failure at full speed is catastrophic; pathogen aerosols must be contained (sealed buckets, opening inside a biosafety cabinet) under DBT/RCGM biosafety guidelines. Laboratory calibration and maintenance records are mandatory under NABL accreditation to ISO/IEC 17025:2017.

Types of centrifugal separation

Differential (pellet) centrifugation: a series of increasing g-force steps; the largest particles pellet first and the supernatant is re-spun. Classical cell fractionation of liver homogenate: 600-1,000 × g/10 min → nuclei and unbroken cells; 10,000-12,000 × g/15 min → mitochondria (plus lysosomes, peroxisomes); 15,000-20,000 × g → lysosomes/peroxisomes; 100,000 × g/60 min → microsomes (ER, plasma-membrane fragments, ribosomes); final supernatant = cytosol. Cheap and fast, but resolution is poor because contamination by trapped and overlapping particles is unavoidable; purity is judged by marker enzymes (succinate dehydrogenase or cytochrome c oxidase for mitochondria, acid phosphatase for lysosomes, catalase for peroxisomes, glucose-6-phosphatase for ER, lactate dehydrogenase for cytosol).
Rate-zonal (density-gradient velocity) centrifugation: sample layered on a pre-formed shallow gradient (usually 5-20% or 15-40% sucrose) and spun for a limited time; separation is by s value, i.e. by size and shape. The run must be stopped before any zone reaches the bottom. Used for ribosome and polysome profiles, subcellular organelles, virus particles.
Isopycnic (equilibrium) centrifugation: long spin in a steep or self-forming gradient; each species bands where gradient density equals its own buoyant density, independent of size. CsCl (ρ ≈ 1.70 g/cm³) separates DNA by GC content and separates supercoiled plasmid DNA from chromosomal DNA with ethidium bromide; sucrose, Percoll, Nycodenz and Ficoll are used for cells and organelles. The Meselson-Stahl experiment (¹⁵N DNA in CsCl) proved semi-conservative replication.
Density-gradient media: sucrose (cheap, hyperosmotic, viscous), Percoll (silica colloid, iso-osmotic, self-forming - lymphocyte and sperm separation), Ficoll-Hypaque (PBMC isolation from blood), CsCl (nucleic acids only).
Analytical ultracentrifugation: small volumes in optical cells with absorbance/interference detection.
- Sedimentation velocity gives s, and with diffusion gives molecular mass, shape (frictional ratio f/f₀) and purity/aggregation.
- Sedimentation equilibrium gives molecular mass independent of shape, and association constants.
Related: differential centrifugal elutriation (cell-cycle synchrony); ultrafiltration spin concentrators (centrifugal, but separation is by membrane cut-off, not sedimentation).

Principle of spectrophotometry

A molecule absorbs a photon when the photon energy matches an electronic transition (E = hc/λ): UV-Vis absorption involves π→π*, n→π* and charge-transfer transitions in chromophores.
Useful regions: far/vacuum UV < 190 nm; UV 200-400 nm; visible 400-800 nm; near-IR beyond. Protein absorbs at 280 nm (Trp, Tyr, and weakly Cys-Cys) and at 190-220 nm (peptide bond); nucleic acids at 260 nm (purine/pyrimidine rings); haem proteins show a Soret band near 410 nm.
Transmittance T = I/I₀; absorbance (optical density) A = log₁₀(I₀/I) = −log₁₀T. A is dimensionless and additive for independent absorbing species (Beer's additivity); %T is not additive, which is why A is used for calculation.
Beer-Lambert law: A = εcl. ε is the molar extinction (absorptivity) coefficient in M⁻¹cm⁻¹ at a stated wavelength, temperature and solvent; it is a constant of the substance.
Measurements are made at λmax because sensitivity is greatest and the spectrum is flattest there, so small wavelength errors matter least.

Instrumentation

Light sources: tungsten-halogen lamp for ~320-1,100 nm; deuterium (or hydrogen) discharge lamp for ~190-400 nm; xenon flash lamp in diode-array instruments; mercury lamp for wavelength calibration.
Monochromator: entrance slit → diffraction grating (or prism) → exit slit; bandwidth (spectral slit width) must be much narrower than the absorption band, else peaks flatten and Beer's law fails.
Sample compartment and cuvettes: optical glass and disposable polystyrene are opaque below ~320 nm, so quartz (silica) cuvettes are compulsory for UV work; standard path length 1 cm, volume 1-3 mL (semi-micro 0.5 mL); cuvettes must be matched, clean, filled above the beam and wiped on the optical faces only.
Detectors: photomultiplier tube (high sensitivity), silicon photodiode, or a photodiode array / CCD in instruments that capture a whole spectrum in one shot.
Designs: single-beam colorimeter with filters (cheap, visible only); single-beam spectrophotometer (zero on a blank, then read); double-beam (beam split between sample and reference - corrects for lamp drift, best for scanning spectra); split-beam diode array; microplate reader (96/384 wells, path length set by the fill volume, so a path-length correction or a standard curve on the same plate is needed); NanoDrop-type micro-volume instruments (1-2 µL, very short path length computed by the instrument).
Related techniques (distinct from absorbance): turbidimetry and nephelometry (scattering, e.g. OD₆₀₀ of bacterial cultures, which is scatter not true absorption); fluorimetry (emission at 90°, 10³ times more sensitive); flame photometry and AAS for metals; spectrofluorimeter, CD and ORD for conformation.
Calibration and QC: wavelength accuracy with holmium oxide glass/solution or a mercury lamp line; photometric accuracy and linearity with potassium dichromate in 0.005 M H₂SO₄ (bands near 235, 257, 313 and 350 nm); stray light with 1.2% KCl at 200 nm (or NaNO₂, NaI); resolution with a 0.02% toluene-in-hexane ratio test. These checks are prescribed in the Indian Pharmacopoeia (IP, current edition 2022) general chapter on UV-Visible spectrophotometry and mirrored in BP/USP; water-quality spectrophotometric methods are specified in IS 3025 (method-wise parts) and CPCB/APHA protocols.

Quantitative applications and assays

Protein by A280: fast, non-destructive, needs no reagent, but needs aromatic residues and is ruined by nucleic acid contamination. Specific absorbance (A280 of a 1 mg/mL solution, 1 cm): BSA ≈ 0.66, IgG ≈ 1.4, lysozyme ≈ 2.64; ε can be computed from sequence as ε₂₈₀ ≈ 5,500·nTrp + 1,490·nTyr + 125·n(S-S).
Warburg-Christian correction estimates protein in the presence of nucleic acid from A280 and A260.
Colorimetric protein assays:

Assay	Chemistry	λ read	Range	Main interference
Biuret	Cu²⁺-peptide bond complex	540-560 nm	1-10 mg/mL	Low sensitivity, ammonium salts
Lowry (Folin)	Biuret + phosphomolybdate reduction	660 or 750 nm	10-300 µg/mL	Detergents, reducing agents, sucrose; non-linear
Bradford	Coomassie G-250 dye binding (red→blue)	595 nm	20-2,000 µg/mL	SDS and detergents; Arg/Lys-biased
BCA	Cu⁺ + bicinchoninic acid (purple)	562 nm	0.5-1,500 µg/mL	Reducing sugars, DTT; detergent-tolerant
UV A280	Trp/Tyr absorption	280 nm	0.1-5 mg/mL	Nucleic acids, buffer absorbance

Nucleic acids: A260 = 1 corresponds to ≈ 50 µg/mL dsDNA, ≈ 40 µg/mL RNA, ≈ 33 µg/mL ssDNA/oligos. A260/A280 ≈ 1.8 for clean DNA and ≈ 2.0 for RNA (lower means protein or phenol); A260/A230 or A260/A270 flags phenol/guanidine. Hyperchromic effect: A260 rises ~40% on denaturation, giving the melting temperature Tm from a thermal melt curve.
Enzyme assays (continuous, kinetic): follow ΔA/min. NADH/NADPH at 340 nm, ε = 6,220 M⁻¹cm⁻¹; p-nitrophenol at 405-410 nm, ε ≈ 18,000 M⁻¹cm⁻¹ (ALP, β-galactosidase); DTNB/TNB⁻ at 412 nm, ε = 13,600; cytochrome c at 550 nm. One International Unit (U) = 1 µmol substrate converted per minute at defined conditions; 1 katal = 1 mol/s.
Other standard colorimetric methods: DNS and Nelson-Somogyi for reducing sugars (540, 520 nm); anthrone/phenol-sulphuric for total carbohydrate (620, 490 nm); GOD-POD for glucose (505 nm); Jaffe for creatinine (520 nm); diazo method for bilirubin; TBA (532 nm) for lipid peroxidation; Nessler and phenate for ammonia in water testing.
Limitations and deviations from Beer's law:
- Real (chemical): association/dissociation, pH-dependent ionisation, complex formation, solvent effects.
- Instrumental: polychromatic light with a wide slit, stray light (the main cause of flattening above A ≈ 2), non-linear detector response.
- Physical: turbidity and scattering, fluorescence of the sample, high concentration (above ~0.01 M the refractive index and ion interactions change ε).
- Reliable working range is roughly A = 0.1-1.0 (best precision near 0.3-0.8); dilute and re-read above that, never extrapolate a standard curve.

Key points

RCF - the centrifugal field as a multiple of g; always quote g and time, not rpm alone.
Stokes' law - sedimentation rate ∝ d² and ∝ (ρp − ρm), inversely ∝ viscosity.
Svedberg unit - 1 S = 10⁻¹³ s; s values are not additive (30S + 50S = 70S).
k-factor - rotor clearing factor; pelleting time in hours = k/s.
Fixed-angle rotor - short path, best for pelleting; swinging bucket - best resolution for gradients.
Differential centrifugation - separation by increasing g steps; purity checked with marker enzymes.
Rate-zonal - separation by s (size/shape); run must be stopped before bands reach the bottom.
Isopycnic - separation by buoyant density only; CsCl for DNA, Percoll for cells.
Analytical ultracentrifugation - sedimentation velocity gives s and shape; equilibrium gives mass.
Beer-Lambert law - A = εcl; valid for monochromatic light and dilute, non-scattering solutions.
Absorbance is additive, transmittance is not - A = −log₁₀T, and A has no units.
Molar extinction coefficient - absorbance of a 1 M solution in a 1 cm path at a stated λ; a property of the substance.
Quartz below 320 nm - glass and plastic cuvettes cannot be used for UV readings.
Double-beam design - compensates lamp and electronic drift; essential for spectral scans.
λmax - chosen for maximum sensitivity and minimum error from wavelength drift.
Working range A 0.1-1.0 - stray light flattens high absorbances; noise dominates low ones.
A260/A280 - 1.8 for DNA, 2.0 for RNA; A260 = 1 → 50 µg/mL dsDNA.
NADH at 340 nm, ε = 6,220 M⁻¹cm⁻¹ - the workhorse of dehydrogenase assays; 1 U = 1 µmol/min.

Formulas and examples

RCF = ω²r/g, with ω = 2πN/60 rad/s, r in m, g = 9.81 m/s²
RCF (× g) = 1.118 × 10⁻⁵ × r(cm) × N²(rpm) — the working form
Stokes: v = d²(ρp − ρm)·ω²r / (18η) [d m, ρ kg/m³, η Pa·s, v m/s]
s = v/(ω²r) (seconds); 1 Svedberg = 10⁻¹³ s
k = ln(rmax/rmin) × 10¹³ / (3600 ω²); pelleting time t(h) = k/s(S)
T = I/I₀; %T = 100 I/I₀; A = log₁₀(I₀/I) = −log₁₀T = 2 − log₁₀(%T)
A = ε c l (ε M⁻¹cm⁻¹, c mol/L, l cm); A = a c l if c is in g/L (a = ε/M)
Protein (mg/mL) ≈ 1.55 A280 − 0.76 A260 (Warburg-Christian)
Enzyme activity (U/mL) = (ΔA/min × V_total) / (ε × l × V_enzyme) with V in mL and ε in mM⁻¹cm⁻¹ ... or use ΔA/min ÷ (ε·l) for mol L⁻¹ min⁻¹

1. rpm to g, and back. A fixed-angle rotor with rmax = 10.0 cm is run at 10,000 rpm.

RCF = 1.118 × 10⁻⁵ × 10.0 × (10,000)² = 1.118 × 10⁻⁵ × 10 × 10⁸ = 11,180 × g.
To obtain 20,000 × g in the same rotor: N² = 20,000/(1.118 × 10⁻⁵ × 10) = 1.789 × 10⁸, so N = 13,375 rpm.
In a microcentrifuge with r = 8.5 cm at 13,000 rpm: 1.118 × 10⁻⁵ × 8.5 × 1.69 × 10⁸ = 16,060 × g — the same rpm, a different field.

2. k-factor and pelleting time. Swinging-bucket rotor, rmin = 5.0 cm, rmax = 10.8 cm, maximum speed 40,000 rpm.

ω = 2π × 40,000/60 = 4,188.8 rad/s; ω² = 1.7546 × 10⁷ s⁻².
ln(10.8/5.0) = ln 2.16 = 0.7701.
k = 0.7701 × 10¹³/(3600 × 1.7546 × 10⁷) = 0.7701 × 10¹³/6.317 × 10¹⁰ = 122.
A 20S particle: t = 122/20 = 6.1 h. An 80S ribosome: t = 122/80 ≈ 1.5 h. At half speed (20,000 rpm) k quadruples to 488, so the 80S run would take ~6.1 h.

3. Transmittance to absorbance and concentration. A coloured complex transmits 25% of the incident light in a 1 cm cell, and ε = 15,500 M⁻¹cm⁻¹.

A = 2 − log₁₀25 = 2 − 1.3979 = 0.602.
c = A/(εl) = 0.602/15,500 = 3.88 × 10⁻⁵ M = 38.8 µM.
If the same solution is diluted 1:1, A = 0.301 and %T = 10^2−0.301 = 50% — absorbance halves, transmittance does not double from 25 to 50 linearly for any other dilution, which is why A is the quantity used.

4. Enzyme activity from a 340 nm assay. An LDH assay in a 3.00 mL cuvette (1 cm) with 0.10 mL of enzyme extract gives ΔA₃₄₀ = −0.062 per minute (NADH consumed), ε = 6,220 M⁻¹cm⁻¹.

Rate of NADH loss = 0.062/6,220 = 9.968 × 10⁻⁶ mol L⁻¹ min⁻¹.
In 3.00 mL (0.00300 L): 9.968 × 10⁻⁶ × 0.00300 = 2.99 × 10⁻⁸ mol/min = 0.0299 µmol/min = 0.0299 U in the cuvette.
Activity of the extract = 0.0299/0.10 mL = 0.299 U/mL.
If the extract contains 1.8 mg protein/mL, specific activity = 0.299/1.8 = 0.166 U/mg.

5. DNA quantitation and purity. A genomic DNA prep diluted 50-fold reads A260 = 0.420 and A280 = 0.220 in a 1 cm quartz cell.

Concentration = 0.420 × 50 µg/mL per A260 unit × 50 (dilution) = 1,050 µg/mL = 1.05 mg/mL.
A260/A280 = 0.420/0.220 = 1.91 — slightly above 1.8, acceptable DNA with a trace of RNA.
Total yield from 2.0 mL of eluate = 1.05 × 2.0 = 2.1 mg.

	Differential	Rate-zonal	Isopycnic
Separates by	Size (s), crudely	s value (size and shape)	Buoyant density only
Medium	Homogenisation buffer	Shallow pre-formed gradient	Steep or self-forming gradient
Run time	Short, fixed steps	Must stop before bands reach bottom	Run to equilibrium
Rotor	Fixed-angle	Swinging-bucket	Vertical / fixed-angle
Typical use	Organelle fractionation	Ribosomes, polysomes, viruses	Plasmid DNA in CsCl, cells in Percoll

Where it is used

Pharma and biotech QC/QA: assay, content uniformity, dissolution and identity tests by UV absorbance per IP/BP monographs at Cipla, Sun Pharma, Dr. Reddy's, Lupin; centrifugal harvest and clarification in vaccine manufacture at Serum Institute of India and Bharat Biotech.
Clinical diagnostics: serum/plasma separation at ~1,500-2,000 × g, Ficoll separation of PBMCs, and semi/fully automated colorimetric analysers for glucose, urea, creatinine, bilirubin, liver and lipid panels — the daily work of lab technologists at SRL, Dr. Lal PathLabs, Metropolis and Thyrocare, under NABL ISO 15189 accreditation.
Research and academia: subcellular fractionation, ribosome profiling, plasmid and virus purification, nucleic-acid and protein quantitation, enzyme kinetics in university, IISc, IITs, CSIR (CCMB, IICT, NCL) and ICAR laboratories.
Food, dairy and water testing: milk-fat testing and cream separation, fruit-juice clarification, spectrophotometric determination of nitrate, nitrite, phosphate, fluoride, iron and residual chlorine by IS 3025 methods in FSSAI-notified and state pollution-board laboratories.
Job roles: QC chemist/analyst, research associate, downstream-processing executive, lab technologist, application specialist and service engineer for instrument companies (Shimadzu, Agilent, Thermo, PerkinElmer, Eppendorf, REMI, Thermo-Eppendorf dealers) — where calibration, 21 CFR Part 11 / ALCOA+ data integrity and IQ/OQ/PQ qualification of a spectrophotometer are routine duties.

Interview questions

Q: Why is sedimentation expressed in g rather than rpm? A: The field depends on radius as well as speed, so the same rpm gives different forces in different rotors; g makes a protocol reproducible on any centrifuge.
Q: Why are 30S and 50S subunits together called 70S and not 80S? A: The sedimentation coefficient depends on mass, shape and hydration, not mass alone, so s values are not additive.
Q: When would you use a swinging-bucket rotor instead of a fixed-angle one? A: For density-gradient separations, because the tube is horizontal during the run so zones stay flat and the longer path gives better resolution.
Q: Difference between rate-zonal and isopycnic centrifugation? A: Rate-zonal separates by sedimentation rate and must be stopped before bands reach the bottom; isopycnic runs to equilibrium and separates purely by buoyant density.
Q: How would you check that your mitochondrial fraction is pure? A: Assay marker enzymes — succinate dehydrogenase or cytochrome c oxidase for mitochondria, acid phosphatase, catalase and glucose-6-phosphatase for lysosomal, peroxisomal and ER contamination.
Q: State the Beer-Lambert law and its limitations. A: A = εcl; it fails at high concentration, with polychromatic light or a wide slit, with stray light, and with turbid, scattering or fluorescent samples.
Q: Why is absorbance preferred to transmittance for calculations? A: Absorbance is linear in concentration and path length and additive across species, whereas transmittance is logarithmic and non-additive.
Q: Why must a quartz cuvette be used at 280 nm? A: Glass and polystyrene absorb strongly below about 320 nm, so they would swamp the sample signal.
Q: Your sample reads A = 2.6. What do you do? A: Dilute and re-read into the 0.1-1.0 range — above about 2 the reading is dominated by stray light and the response is no longer linear.
Q: An A260/A280 ratio of 1.5 for a DNA prep means what? A: Protein or phenol contamination; re-extract or re-purify before using the DNA in enzymatic reactions.
Q: How do you convert ΔA/min at 340 nm into units of enzyme activity? A: Divide ΔA/min by ε (6,220 M⁻¹cm⁻¹) and the path length to get mol L⁻¹ min⁻¹, multiply by the assay volume in litres to get mol/min, express as µmol/min (= U) and divide by the volume of enzyme added.
Q: Which checks establish that a UV-Vis spectrophotometer is fit for pharmacopoeial use? A: Wavelength accuracy with holmium oxide or a mercury lamp, photometric accuracy and linearity with potassium dichromate solutions, stray light with KCl or NaNO₂, and resolution with the toluene-in-hexane ratio, as specified in the Indian Pharmacopoeia.

Common mistakes

Quoting rpm instead of × g, or recalculating g with the wrong radius (use rmax for pelleting, rav for comparisons — and state which).
Forgetting that a rotor must be derated for high-density media and for age, and running unbalanced tubes.
Treating s values as additive, or expecting isopycnic banding to depend on particle size.
Reading absorbances above ~1.5-2 and fitting them on a standard curve, or extrapolating beyond the highest standard.
Using a glass or plastic cuvette for 260/280 nm readings, or blanking against water when the assay reagent itself absorbs.
Reporting protein from A280 on a nucleic-acid-contaminated lysate without the Warburg-Christian correction, or applying BSA-based Bradford standards to a protein with very different Arg/Lys content.
Mixing units in Beer-Lambert — ε in M⁻¹cm⁻¹ with concentration in mg/mL, or forgetting the dilution factor and the assay-to-enzyme volume ratio in activity calculations.