Analytical & QC - Mika Biologics

Analytical & QC CDMO Serivces for Microbial Biologics

Design → Data → Decision. When the product is microbial-first—phage, lysin, bacteriocin, cytokine from E. coli, VLP/OMV/EV, Pichia-expressed biobetter, or an engineered probiotic—analytics must read the product’s truth without distortion. Mika Biologics builds analytical systems that survive validation, anticipate audits, and shorten decision-time.

Executive overview

Microbial biologics do not behave like mammalian proteins, and vesicles do not behave like soluble proteins. Phage titers can appear “healthy” while residual DNA drifts; lysins stay pure but lose the very activity that justifies the programme; bacteriocins purify beautifully but carry the wrong salt legacy; VLP/OMV/EV particles measure one size by DLS and quite another by NTA; Pichia-derived antibodies look perfect on SEC but tell a different story under glycopeptide LC–MS; engineered probiotics pass CFU yet fail payload potency. The common failure mode is not manufacturing; it is measurement that cannot see what matters, or cannot see it the same way twice.

Our Analytical & QC platform is built specifically for microbial programmes. We start with QTPP → CQAs → CPPs, define the control strategy in plain terms, and then assemble method stacks that are phase-appropriate, scalable, and inspector-proof. The emphasis is not on the most exotic instrument; it is on convergence—orthogonal methods that tell the same story from different angles. Results are captured in our LIMS/eBatch backbone with ALCOA+ discipline, surfaced in dashboards that link raw data → trend → decision, and packaged into COAs reviewers can navigate in minutes.

This page goes deep. It includes method menus, sample COAs, particle analytics for VLPs/EVs, rFC/LAL endotoxin strategy, bioassays, EM/DLS/NTA convergence, method-transfer SLAs, and the digital quality spine that keeps everything inspection-ready.

Scope: what we test

Identity & integrity — peptide mapping, intact mass, genome identity (phage), glycan fingerprints (Pichia), vesicle markers (CD9/63/81; OMP panels).
Potency — PFU/EOP (phage), kinetic lysis for lysins, AU/MIC/time-kill for bacteriocins, enzymatic units for payloads, receptor binding (SPR/BLI), ADCC/CDC for Fc constructs, functional vesicle assays, probiotic payload activity.
Purity & impurities — SEC-MALS aggregates, HCP/HCD, residual DNA (qPCR/ddPCR), LPS/endotoxin (LAL/rFC), host lipids and carbohydrates (β-glucan, mannan), particle co-contaminants in vesicle preps.
Particles & morphology — DLS, NTA, EM (negative stain or cryo), AUC (as needed), subvisible counts and morphology by MFI for protein products.
Safety & sterility — sterility (pharmacopoeial), bioburden, adventitious agents (phage screen when relevant), AR gene surveillance for LBPs.
Physicochemical & stability — icIEF/cIEF, CE-SDS, HIC/IEX/RP-HPLC maps, pH/osmolality/viscosity; ICH stability with drift indicators tied to mechanism (e.g., sialic acid loss, phage titer decay, vesicle aggregation, CFU decay).
Comparability & lifecycle — method validation/qualification per ICH, pre-agreed comparability bands, PPQ acceptance, CPV trend logic.

Method catalogue

Identity & structure

LC–MS/MS peptide mapping with coverage metrics; intact mass for DS/DP identity.
Glycan analytics (released HILIC-UPLC-FLR with ESI-MS; exoglycosidase confirmation; linkage-specific derivatisation for α2,3/α2,6; glycopeptide LC–MS/MS for site occupancy).
Genomic identity for phage (whole-genome sequencing, termini/packaging site calls), engineered probiotics (strain-specific markers + periodic WGS snapshot).
Vesicle markers: exosomes (CD9/CD63/CD81, TSG101/ALIX positive; GM130/calnexin negative), OMVs (defined OMPs; lipid A profile where applicable).

Potency & function

Phage: PFU/mL by validated double-layer agar, EOP on isolate panels, liquid-kill kinetics.
Lysins: turbidimetric kinetic lysis (Target Kill Units), MIC/time-kill, zymograms.
Bacteriocins/lantibiotics: AU/mL (agar diffusion/turbidity), MIC/time-kill, lipid II-proxy leakage assays for mechanism context.
Cytokines/enzymes: enzyme activity per mg; cell-based potency (e.g., STAT reporter for cytokines).
Fc constructs: ADCC reporter, CDC, SPR/BLI (FcγRIIIa, FcRn).
VLP/OMV/EV: antigen exposure panels, immune activation surrogates, cell uptake/functional shifts.
Engineered probiotics: CFU-normalised payload activity (enzyme, cytokine); colon-release performance; environmental challenge response for kill-switches.

Purity, impurities, and safety

SEC-MALS (aggregate mass fractions), DLS (hydrodynamic radius), AUC (as needed for difficult oligomers).
HCP/HCD: platform ELISAs for E. coli and yeast/fungi; orthogonal LC-MS peptide traces when required.
Residual DNA: qPCR/ddPCR with host-specific assays; phage DNA residuals where relevant.
Endotoxin: LAL per pharmacopoeia and recombinant Factor C (rFC) per USP <86>—full method suitability, interference controls, spike recovery, and acceptance criteria tied to route/dose.
Sterility/bioburden: compendial methods with suitability; closed-system bioburden trending for aseptic steps.
Host carbohydrates & lipids: β-glucan/mannan (yeast matrices), lipid carry-over (OMV/EV), detergent/solvent residuals (if used in DSP).

Particle analytics & morphology (VLP/OMV/EV emphasis)

NTA for particle concentration and size distribution—calibrated; reportable range locked; viscosity corrections.
DLS for hydrodynamic profiles—used as an orthogonal view to NTA; not a replacement.
EM: negative-stain for development speed; cryo-EM where morphology at near-native state is pivotal.
SEC-MALS to separate and weigh assemblies; AUC for complex multimers; MFI when protein particles are the CQA (e.g., protein DP).

Physicochemical

icIEF/cIEF for charge variants; CE-SDS (reducing/non-reducing); HIC/IEX/RP-HPLC maps.
pH/osmolality/viscosity; moisture & water activity for lyo; headspace O₂ when dosage form demands.

Microbial viability & function (LBP focus)

CFU with strain-specific plating and diluent conditions (anaerobic vs aerobic).
Fast viability by flow cytometry (membrane integrity dyes) for in-process decisions.
Genetic stability metrics (plasmid loss, integration integrity, reversion rates), kill-switch performance.

How we design analytical control: QTPP → CQAs → CPPs

We begin each programme with a one-page translation of intent:

QTPP (examples): route (IV, SC, oral), dose, presentation (vial, PFS, capsule/sachet), storage (2–8 °C or ambient lyo), shelf-life target, potency mechanism (ADCC, enzymatic rate, PFU, particle function), particle spec (if vesicles).
CQAs (examples): identity (LC–MS or genome), potency (mechanism-true assays), purity (HCP/HCD/residual DNA), endotoxin, aggregation, particle size/count, glycan states (fucose/sialylation/bisecting), CFU and payload activity (LBP), sterility/bioburden, appearance/pH/osmolality.
CPPs: the few levers that actually move CQAs—e.g., nuclease exposure and AEX set-points for phage residual DNA/LPS, shear and residence time for vesicle size, redox and feed composition for Pichia glycan occupancy, ORP and headspace O₂ for LBP viability.

Analytics then mirror this map, with acceptance bands agreed early and statistical rules that discriminate real movement from instrument noise.

Method transfer SLAs

You get a documented pathway from sponsor method to Mika method:

Pre-transfer alignment — receive SOPs, raw data exemplars, system suitability templates, and critical reagents list; identify gaps (columns, ligands, antibodies, reference standards).
Bridging experiments — side-by-side runs on sponsor vs Mika instruments where practicable; predefined criteria for bias (slope/intercept), precision, and range.
Report & lock — method adaptation report with any justified deviations; updated SOPs; analyst training records; data integrity checks.
Qualification/validation — phase-appropriate: specificity, accuracy, precision, linearity, range, robustness, LOD/LOQ where required; we follow ICH expectations without inflating effort.
SLA windows (typical) — single-assay transfers complete in a short, pre-agreed window; multi-assay panels sequenced with milestone gates; urgent lot-release methods accommodated with explicit risk posture and temporary controls.

We state what success looks like, what is out-of-scope, and who owns which reagent up front. No ambiguity, no surprise invoices.

A) Phage Drug Product — Vial (liquid)

Identity

Genome identity: confirmed (WGS; no lysogeny/toxins/ARGs)
Particle morphology: EM consistent with lytic family

Potency

Titer (PFU/mL): 1.2 × 10¹⁰ (Spec: ≥1.0 × 10¹⁰)
EOP on panel: within predefined matrix (Spec: pass)

Purity & safety

Endotoxin (EU/mL): 0.18 (Spec: ≤0.5)
Residual host DNA: 6.0 ng/mL (Spec: ≤10 ng/mL)
Residual host proteins: < Spec threshold
Sterility: pass
Appearance: clear, colourless
pH/osmolality: in range

Stability

Timepoint: Month 3 (2–8 °C) potency ≥95% of T0

B) Lysin Enzyme — DP (lyo)

Identity

LC–MS intact mass: matches theoretical; peptide map ≥95% coverage

Potency

Turbidimetric lysis: ≥8,000 TKU/mg (Spec: ≥7,500)
MIC (target): within acceptance

Purity & safety

Purity (UHPLC): ≥98%
Aggregates (SEC-MALS): ≤1.5%
Endotoxin (EU/mg): ≤0.5
Sterility: pass
Residual DNA: ≤10 ng/dose
Appearance: off-white cake; reconstitution ≤2 min

C) Bacteriocin (Nisin) — Ingredient-grade

Identity

LC–MS peptides: consistent with variant A

Potency

Activity: ≥1,000 IU/mg (Spec: ≥900)

Purity

HPLC purity: ≥95%
Moisture: ≤4%
Microbiology: TPC/Y&M within limits

D) VLP — Vaccine DP (liquid)

Identity & morphology

NTA: 2.5 × 10¹¹ particles/mL (Spec: ≥2.0 × 10¹¹)
Size (mode): 95 nm (Spec: 80–120 nm)
EM: intact particles; expected symmetry

Potency

Antigen display ELISA: ≥1.0 AU/µg
Cell-based activation: pass

Purity & safety

Protein impurities: ≤ Spec
Endotoxin: ≤ Spec (route-dependent)
Sterility: pass
Aggregates (SEC-MALS): ≤5% > 200 nm

E) Exosomes/OMVs — DP (liquid or frozen)

Identity

Markers: CD9/CD63/CD81 positive; GM130 negative (exosomes)
OMP panel (OMVs): within profile

Particles

NTA: ≥X particles/mL; size window met
DLS PDI: ≤0.25

Safety

Endotoxin: ≤ route-specific limit
Sterility: pass

Potency

Functional assay: in acceptance band

F) Pichia-derived Fc Biobetter — DS (liquid)

Identity & glycofingerprint

Released-glycan: afucose ≤1.0%; α2,6-Sia ≥12%; bisecting ≥15%
Glycopeptide occupancy: site-specific bands met

Function

FcγRIIIa SPR: KD in band
ADCC reporter: ≥ Spec

Purity & safety

SEC-MALS aggregates: ≤1.0%
HCP (yeast): ≤ Spec
Endotoxin: ≤ Spec
Residual DNA: ≤ Spec

G) Engineered Probiotic — DP (lyo capsule)

Identity

Strain confirmed (PCR/WGS snapshot)

Viability & potency

CFU/capsule: ≥1.0 × 10¹⁰ (Spec: ≥1.0 × 10¹⁰)
Payload activity: ≥ Spec (enzyme Units/capsule)

Stability & safety

Residual moisture: 1.8% (Spec: ≤3.0%)
Package O₂: ≤1.0% (Spec: ≤2.0%)
Environmental challenge kill-switch: pass
Pathogen panel: negative

Particle analytics for VLPs/EVs: convergence as policy

NTA gives concentration and a size distribution with per-bin counts; it is sensitive to settings and viscosity. DLS reports an intensity-weighted hydrodynamic size that over-weights larger particles; it is fast and repeatable with tight SOPs. EM sees morphology and integrity. We do not pick one—we calibrate and reconcile all three so reviewers see the same particle from complementary vantage points. SEC-MALS then links the ensemble back to mass and purity. When the story converges, batch release gets simpler and comparability survives scrutiny.

Endotoxin strategy: rFC/LAL without myths

Endotoxin is designed out where possible (host/media choice, AEX/TFF polishing) and tested with methods that make scientific and ethical sense. We qualify LAL per compendia and, where appropriate, deploy recombinant Factor C (rFC) per USP guidance. Method suitability is non-negotiable: spike recovery (inhibition/enhancement), dilution policies, and masking agents if chemically justified. We set acceptance criteria by route/dose and narrative—protein injectables follow familiar limits; OMV/EV programmes set limits by risk and claim; oral LBPs justify limits scientifically.

Bioassays that reward good process

A bioassay that cannot discriminate “good” from “almost good” wastes time and money. We choose mechanism-true assays (ADCC, enzymatic rate in relevant matrix, PFU/EOP, antigen display + activation, payload activity at gut-like pH) and make them repeatable: clear controls, reference standards, acceptance bands, and outlier rules. Where possible, we include orthogonal surrogates (e.g., FcγRIIIa binding by SPR alongside ADCC) so release never hangs on a single noisy number.

Data integrity, LIMS, and eBatch: audit without theatre

All results land in our LIMS with ALCOA+ enforcement. Analyst IDs, instrument IDs, audit trails, versioned SOPs/training, attachment of raw files (chromatograms, spectra, micrographs), and automatic trend plots are standard. Batch records link unit operation → sample → method → result; CPV dashboards correlate historian tags (e.g., TFF flux/ΔP, conductivity profiles, lyo Pirani/capacitance) with CQAs (e.g., particle size drift, aggregate content). Remote review is fast because the story is already coherent.

Stability to ICH: conditions that reveal, not invent, movement

We run long-term and accelerated arms with stress designed to mirror real-world hazards (freeze–thaw, agitation, light). Each programme has stability-indicating methods tied to mechanism:

Phage: PFU/M and residual DNA drift.
Lysin/bacteriocin: potency, aggregates, oxidation/deamidation.
VLP/EV/OMV: particle size/count, morphology, endotoxin drift, potency.
Pichia Fc: glycan drift (sialic acid loss), aggregates, charge variants, potency.
LBPs: CFU decay, payload activity retention, package O₂/moisture creep, genetic stability.

We define interim actions (what you do if Month-3 potency drops 7%) and connect stability to label and logistics.

Comparability, PPQ & CPV

Comparability plans state exactly which assays and bands carry the argument when you change scale, site, or process. We avoid “over-measuring” and focus on the markers that predict performance. PPQ closes the loop with acceptance criteria tied to real process variability, not aspirational specs. CPV then watches the few historian signals that truly correlate with CQAs and raises alerts before release results move.

Why Mika for microbial analytics

Microbial-first coverage across phage, enzymes, bacteriocins, VLP/OMV/EV, Pichia biobetters, and engineered probiotics.
Orthogonal design so no single method owns the truth.
Method transfer without drama—SLAs, bridging, and validation that match phase.
Digital QMS that makes audits short and pleasant.
Interlocking pages & teams—Formulation & Fill-Finish for CCIT/lyo, Process Characterisation for PPQ/CPV, and service lines that share controls (e.g., endotoxin logic across modalities).

Inter-page guidance

For particle products, see Exosomes & OMVs — Bioprocessing for upstream/downstream context.
For glyco-tuned proteins, see Advanced Yeast Glycoengineering (Pichia) for Biobetters for how analytics steer design.
For LBPs, see Engineered Probiotics (LBP) — GMP Manufacturing for chassis, kill-switches, and dosage forms.
For fill-finish, see Formulation & Aseptic Fill-Finish (Grade A Isolators).
For lifecycle, see Process Characterisation, PPQ & CPV for Microbial Platforms.

Analytical & QC FAQ

Q1. Can you run rFC instead of LAL for endotoxin?
Yes. We qualify recombinant Factor C per compendial guidance where appropriate. Choice of rFC vs LAL is made during method suitability with spike-recovery evidence and justified in the dossier.

Q2. How do you stop NTA and DLS from contradicting each other?
We don’t force agreement; we calibrate and interpret. NTA carries number-based distributions; DLS carries intensity-weighted hydrodynamics. We standardise settings, viscosity, and dilution, then add EM and SEC-MALS so convergence is demonstrable.

Q3. Our sponsor method uses a specific column or antibody that is hard to source. Can you still transfer?
Yes. We document the supply risk, propose qualified alternates, and run bridging experiments. If the method must remain as-is, we maintain a controlled inventory with expiry tracking.

Q4. Can you release phage products where 0.22 µm filtration is not possible?
Yes. We release on aseptic closed processing with sterility tests, bioburden controls, and environmental monitoring, plus endotoxin/residuals within limits. The analytical package is written to explain why filtration is inappropriate for the product.

Q5. What does method validation look like for early phase?
Phase-appropriate: specificity and precision are non-negotiable; accuracy/linearity/range/robustness are qualified to the degree required to make safe decisions. Full validation follows as the programme matures.

Q6. Do you offer ADCC/CDC and receptor binding together?
Yes. We prefer mechanism-true ADCC/CDC paired with SPR/BLI receptor kinetics; the former proves function, the latter stabilises release when bioassay noise is inevitable.

Q7. How do you measure residual DNA in phage and protein products?
Assay-specific qPCR/ddPCR with appropriate standards and controls; for phage we also report genome integrity and packaging; for proteins we target host-specific DNA with validated LOD/LOQ.

Q8. We have a defined LBP consortium—can you control per-strain ratios at release?
Yes. We set ratio bands, quantify by qPCR and CFU, and run shipping simulations. Homogeneity and drift are part of the validation story.

Q9. Can you act as “outsourced SEC-MALS/SPR/BLI”?
Yes. SEC-MALS for aggregation, SPR/BLI for binding kinetics, with method transfer options and data packaged into COAs and comparability addenda.

Q10. How do you keep glycan fingerprints stable across lots?
We link process levers to analytics (redox, feed, temperature, secretory rate), lock released-glycan and glycopeptide banding, and tie potency (ADCC, FcRn binding) to those bands. Deviations trigger root cause and comparability pathways.